#!pip install lightgbm

%%time
print('loading imports...')

%matplotlib inline

# Import data and general utilities
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
import statistics as st
import scipy as sp
import sklearn
from tqdm import tqdm

import plotly.graph_objs as go
import plotly.subplots as sp
import ipywidgets as widgets
import math
import statsmodels
import warnings
import io
import inspect
import sys
import datetime
import ipywidgets as widgets
import traceback
import os
from tabulate import tabulate
import pprint

from scipy.stats.mstats import winsorize
from IPython.display import display, HTML
from collections import Counter  
from scipy.stats import skew, kurtosis

# Scaling and preprocessing
from sklearn.preprocessing import RobustScaler, StandardScaler, LabelEncoder
from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LinearRegression
from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error, median_absolute_error, explained_variance_score, mean_squared_log_error
from sklearn.inspection import permutation_importance
import statsmodels.api as sm
from sklearn.linear_model import HuberRegressor
from sklearn.metrics import rand_score
from statsmodels.stats.stattools import durbin_watson
from scipy.stats import zscore
from scipy.stats import median_abs_deviation

# Learning and prediction algorithms (Classifiers)
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.neural_network import MLPClassifier
from sklearn.ensemble import AdaBoostClassifier, GradientBoostingClassifier, ExtraTreesClassifier, BaggingClassifier, StackingClassifier, VotingClassifier, HistGradientBoostingClassifier
from xgboost import XGBRegressor
from sklearn.cluster import DBSCAN

# Learning and prediction algorithms (Regressors)
from sklearn.linear_model import LinearRegression, Ridge, Lasso, SGDRegressor, ElasticNet, LassoCV, RidgeCV, ElasticNetCV, BayesianRidge, ARDRegression, HuberRegressor, RANSACRegressor, TheilSenRegressor
from sklearn.tree import DecisionTreeRegressor
from sklearn.svm import LinearSVR, SVR
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor, AdaBoostRegressor, ExtraTreesRegressor, HistGradientBoostingRegressor 
from xgboost import XGBRegressor 
from sklearn.gaussian_process import GaussianProcessRegressor
from sklearn.neural_network import MLPRegressor 
from sklearn.ensemble import ExtraTreesRegressor 
from sklearn.neighbors import KNeighborsRegressor
from catboost import CatBoostRegressor
from lightgbm import LGBMRegressor

# Statsmodels
from statsmodels.graphics.gofplots import ProbPlot
from statsmodels.formula.api import ols
import statsmodels.api as sm

# Other sklearn utilities
from sklearn import tree
from sklearn.model_selection import GridSearchCV #For grid search

#to ignore warnings
warnings.filterwarnings("ignore")


print('\nimports complete\n')

loading imports...

imports complete

CPU times: user 2.63 s, sys: 247 ms, total: 2.88 s
Wall time: 9.86 s

def get_time():
    formatted_time = ""

    try:
        now = datetime.datetime.now()

        year = now.year
        month = now.month
        day = now.day
        hour = now.hour
        min = now.minute

        formatted_time = "{}-{:02}-{:02}-{:02}{:02}".format(year, month, day, hour, min)

    except Exception as error:
        print_exception(error, inspect.currentframe().f_code.co_name, False)

    return formatted_time

def calculate_time_duration(datetime_duration):
    formatted_time = ""

    try:
        days = datetime_duration.days
        hours = datetime_duration.seconds // 3600
        minutes = (datetime_duration.seconds % 3600) // 60
        # seconds = str(divmod(datetime_duration.seconds, 1))
        seconds = datetime_duration.seconds % 60

        formatted_time = f"{days} day(s), {hours} hour(s), {minutes} minute(s), {seconds} second(s)"

    except Exception as error:
        print_exception(error, inspect.currentframe().f_code.co_name, False)

    return formatted_time

def display_NaN_columns(dataframe):
    try:
        
        lst_nan_sum = dataframe.isna().sum()

        filtered_cols_with_na_sum_greater_than_zero = lst_nan_sum[lst_nan_sum > 0]

        if len(filtered_cols_with_na_sum_greater_than_zero) < 1:
            print("NO COLS WITH NaN!!!!!!!!!!!!")
        else:
            print(f'Columns with NaN: {len(filtered_cols_with_na_sum_greater_than_zero)}')
            print(filtered_cols_with_na_sum_greater_than_zero)
        
        
    except Exception as error:
        print_exception(error, inspect.currentframe().f_code.co_name, False)

def print_nunique(SERIES, num_columns):
    # iterate through df
    if num_columns < 1:
        num_columns = 1
    try:
        count = 0
        for col, vol in SERIES.items():
            print(f'{col}:  {vol}', end='\t\t\t\t')
            count += 1
            
            if count % num_columns == 0:
                print('')
            
    except Exception as error:
        print_exception(error, inspect.currentframe().f_code.co_name, False)

lst_coef_only_models = ['svm_regressor_linear_1', 'linear_regression', 'linear_regressor', 'svm_regressor_linear_1', 'svm_regressor_linear_2', 'lasso_regressor', 'grid_search_regressor']
lst_coef_models = ['logistic_regression', 'perceptron', 'sgd_classifier', 'passive_aggressive', 'ridge_classifier', 'linear_svc', 'svc', 'svc_nu']
lst_all_feature_importance = ['decision_tree_pruned_regressor', 'decision_tree_regressor', 'bagging_classifier', 'random_forest_regressor', 'extra_trees_regressor', 'xgb_regressor_1', 'xgb_regressor_2', 'lgbm_regressor', 'decision_tree_regressor', 'decision_tree_pruned_regressor', 'gradient_boosting_regressor_1', 'gradient_boosting_regressor_2', 'gradient_boosting_regressor_3', 'random_forest_regressor', 'random_forest_regressor_1', 'random_forest_regressor_2', ] 
lst_permutation_importance = ['nbc_gaussian', 'knn_regressor_1', 'knn_regressor_2', 'ada_boost_regressor', 'quadratic_discriminant', 'linear_discriminant', 'hist_gradient_boosting_regressor', 'gaussian_process_regressor', 'mlp_regressor_1', 'mlp_regressor_2', 'nbc_bernoulli', 'ransac_regressor', 'svm_regressor', 'svm_rbf_regressor', 'svm_poly_regressor', 'svm_sigmoid_regressor', 'theil_sen_regressor']
lst_get_feature_importance = ['catboost_regressor',]

def display_feature_importance_chart(model, model_name, lst_feature_names, num_coefficients_to_display, figure_save_name_ok_to_be_null, X_test, y_test, display_chart):

    try: 
        ftr_importance = None;
        
        # not all functions have the same protocol to extract feature importance coefficients
        ####################################################################################
        # model.coef_   --> lst_coef_only_models
        ####################################################################################
        if model_name in lst_coef_only_models:
            try:
                ftr_importance = np.abs(model.coef_)

                if ftr_importance.ndim > 1:
                    ftr_importance = np.mean(np.abs(ftr_importance), axis=0) #Average across the first dimension
            except:
                ftr_importance = np.abs(model.coef_)

        ####################################################################################
        # model.coef_[0] --> lst_coef_models
        ####################################################################################
        elif model_name in lst_coef_models:
            ftr_importance = np.abs(model.coef_[0])

        ####################################################################################
        # model.feature_importances_ --> lst_all_feature_importance
        ####################################################################################
        elif model_name in lst_all_feature_importance:
            if hasattr(model, 'estimators_'): #Added a check to see if the model has the attribute estimators_
                try:
                    all_feature_importances = np.array([tree.feature_importances_ for tree in model.estimators_])
    
                    # Average the feature importances across all trees
                    mean_feature_importances = np.mean(all_feature_importances, axis=0)
    
                    ftr_importance = mean_feature_importances
                except:
                    ftr_importance = model.feature_importances_

            # check if we have importances:
            if ftr_importance is None or len(ftr_importance) < 1:
                ftr_importance = model.feature_importances_
                    
            else: # tree based models
                ftr_importance = model.feature_importances_ #If it does not have the attribute, then it is a single model, so it has feature_importances_ directly
        
        ####################################################################################
        # results.importances_mean --> lst_permutation_importance
        ####################################################################################        
        elif model_name in lst_permutation_importance:
            results = permutation_importance(model, X_test, y_test, scoring='neg_mean_squared_error', n_repeats=10, random_state=42, n_jobs=-1) #Added n_jobs and changed scoring to neg_mean_squared_error for regressors

            # Get feature importances
            importance = results.importances_mean
            ftr_importance = np.abs(importance)

        ####################################################################################
        # get_feature_importance()  --> lst_get_feature_importance
        ####################################################################################
        elif model_name in lst_get_feature_importance:
            ftr_importance = model.get_feature_importance()

            if ftr_importance is None or len(ftr_importance) < 1:
                #You can also specify the type of importance you want.
                ftr_importance = model.get_feature_importance(type="PredictionValuesChange") #Other options include "LossFunctionChange" and "ShapValues"
            
                                             
        ####################################################################################
        # else model.coef_
        ####################################################################################
        else:            
            ftr_importance = np.abs(model.coef_)

        return display_feature_importance(model_name, lst_feature_names, ftr_importance, num_coefficients_to_display, figure_save_name_ok_to_be_null, display_chart)
        
    except Exception as e:
        print(f"* * * Error in display_feature_importance_chart for {model_name}: {e}")
        return None

#display feature importance
def display_feature_importance(model_name, list_col_names, feature_importances, num_features, figure_save_name_ok_to_be_null, display_chart):
    try:
        # make directory
        os.makedirs('./feature_importance', exist_ok=True)
        os.makedirs('./feature_importance_data', exist_ok=True)

        
        #plot importances
        feature_names = list_col_names
        importances = feature_importances
        indices = np.argsort(importances)

        # save coefficients to file
        # reverse indices
        reverse_indices = indices[::-1]
        try:
            if figure_save_name_ok_to_be_null is not None and len(figure_save_name_ok_to_be_null) > 0:
                # Export to a file
                with open(str('./feature_importance_data/' + figure_save_name_ok_to_be_null) + '.csv', 'w') as file:
                    # write header
                    file.write(str(model_name) + '_feature,importance_coefficient\n')
                    
                    # write values
                    for i in reverse_indices:
                        file.write(str(list_col_names[i]) + "," + str(importances[i]) + '\n')
        except:
            pass
        

        # reduce the number of values to print to cell
        if num_features > 0:
            indices = np.argsort(importances)[-num_features:]

        
        figure_height = math.ceil(num_features / 2)
        
        if figure_height < 1:
            figure_height = math.ceil(len(feature_importances) / 3)
        
        plt.figure(figsize=(15, figure_height))
        plt.title("Feature Importances for model [" + str(model_name) + "]")
        plt.barh(range(len(indices)), importances[indices], align="center", edgecolor='black')
        plt.yticks(range(len(indices)), [feature_names[i] for i in indices])
        plt.xlabel("Relative Importance")
        plt.grid(axis='x', linestyle='--', alpha=0.7)

        # ensure all text fits within figure
        plt.tight_layout()
        
        try:
          if figure_save_name_ok_to_be_null is not None and len(figure_save_name_ok_to_be_null) > 0:
              # Save the figure
              plt.savefig('./feature_importance/' + str(figure_save_name_ok_to_be_null) + '.png', bbox_inches='tight')    # add 'tight' to ensure all text fits inside the figure
        except:
            pass
          
               
        dict_coefficients = dict()
        
        if display_chart:        
            #display list of coefficients
            print("coefficients:\n====================")        
            if num_features > 0:
                reverse_indices = indices[::-1]
                
            #print
            for i in reverse_indices:
                print(str(list_col_names[i]) + "\t" + str(importances[i]))
                dict_coefficients[str(list_col_names[i])] = str(importances[i])

            # show chart
            plt.show()

        return dict_coefficients, plt
            
    except Exception as error:
        print(error)
        print(inspect.currentframe().f_code.co_name)

    print('\n')

def plot_boxplot_plotly(df_to_analyze, plot_title, fill_color):
    num_cols = 4
    num_rows = int(np.ceil(len(df_to_analyze.columns) / num_cols))
    
    # Create a subplot grid
    fig = sp.make_subplots(rows=num_rows, cols=num_cols, subplot_titles=df_to_analyze.columns)

    # Iterate through each feature and create a boxplot
    for i, col in enumerate(df_to_analyze.columns):
        row = (i // num_cols) + 1
        col_pos = (i % num_cols) + 1

        if fill_color is None:
            fig.add_trace(go.Box(y=df_to_analyze[col], name=col, boxpoints="outliers"),row=row, col=col_pos)
        else:
            fig.add_trace(go.Box(y=df_to_analyze[col], name=col, boxpoints="outliers", marker_color=fill_color),row=row, col=col_pos)

    # Update layout
    fig.update_layout(
        title_text=plot_title,
        height=500 * num_rows,  # Adjust figure height dynamically
        showlegend=False
    )

    fig.show()

def plot_kde_histplot(df_to_analyze, set_y_as_density, num_cols, plot_title):           

    # set num rows
    num_rows = int(np.ceil(len(df_to_analyze.columns)/num_cols))

    # set y label
    y_label = 'Frequency'
    if set_y_as_density:
        y_label = 'Density'    

    
    fig, axes = plt.subplots(num_rows, num_cols,figsize=(20, 6*num_cols))
    
    fig.suptitle(plot_title,y=0.99, size=20)
    
    axes = axes.flatten()
    
    for i, attr in enumerate(df_to_analyze.columns):
        if set_y_as_density:
            #sns.kdeplot(df_to_analyze[attr], ax=axes[i], fill=True,  alpha=0.7, linewidth=0.75, edgecolor='k')
            sns.kdeplot(df_to_analyze[attr], ax=axes[i], fill=True,  linewidth=0.75)
        else:
            sns.histplot(df_to_analyze[attr], bins=10, ax=axes[i], linewidth=1, kde=True)
            axes[i].set_ylabel(y_label)
            axes[i].grid(True, axis='y', linestyle='--')
            
        SKEW = str(round(df_to_analyze[attr].skew(axis = 0, skipna = True),2))
        KURTOSIS = str(round(df_to_analyze[attr].kurt(axis = 0, skipna = True),2))
        
        axes[i].set_title(f'{attr} - skewness: {SKEW}, kurtosis: {KURTOSIS}'  )
        axes[i].tick_params(axis='both', which='both', labelsize=10)

        # add mean and median line
        mean_attr = df_to_analyze[attr].mean()
        median_attr = df_to_analyze[attr].median()
    
        axes[i].axvline(mean_attr, color='orange', linestyle='--', label=f'Mean: {mean_attr:.2f}')
        axes[i].axvline(median_attr, color='green', linestyle='-', label=f'Median: {median_attr:.2f}')
    
        axes[i].legend()

    # Hide unused subplots if num_cols * num_rows > num features
    for j in range(i + 1, len(axes)):
        fig.delaxes(axes[j])

def display_hist_kde_plot(df_to_analyze, set_y_as_density, num_cols, plot_title):           

    # set num rows
    num_rows = int(np.ceil(len(df_to_analyze.columns)/num_cols))

    # set y label
    y_label = 'Frequency'
    if set_y_as_density:
        y_label = 'Density'    
    
    # uncomment out below if you wish to share x and y data points
    #fig, axes = plt.subplots(num_rows, num_cols, figsize=(15, 8), sharex=True, sharey=True)
    fig, axes = plt.subplots(num_rows, num_cols, figsize=(20, 6*num_cols), sharex=False, sharey=False)
    fig.suptitle(plot_title,y=1, size=20)
    
    # flatten subplot for iteration
    axes = axes.flatten()        
    
    # plot histogram in each subplot
    for i, attr in enumerate(lst_features):
        sns.kdeplot(df_to_analyze[attr], ax=axes[i], color='red', linewidth=2, alpha=0.8)
        axes[i].hist(df_to_analyze[attr], bins=10, edgecolor='black', alpha=0.7, density=set_y_as_density)

        SKEW = str(round(df_to_analyze[attr].skew(axis = 0, skipna = True),2))
        KURTOSIS = str(round(df_to_analyze[attr].kurt(axis = 0, skipna = True),2))
        
        axes[i].set_title(f'{attr} - skewness: {SKEW}, kurtosis: {KURTOSIS}')
        axes[i].set_xlabel(attr)
        axes[i].set_ylabel(y_label)
        axes[i].tick_params(axis='both', which='both', labelsize=10)
    
        # add mean and median line
        mean_attr = df_to_analyze[attr].mean()
        median_attr = df_to_analyze[attr].median()
    
        axes[i].axvline(mean_attr, color='orange', linestyle='--', label=f'Mean: {mean_attr:.2f}')
        axes[i].axvline(median_attr, color='green', linestyle='-', label=f'Median: {median_attr:.2f}')
    
        axes[i].legend()
        axes[i].grid(True, axis='y', linestyle='--')
    
    # Hide unused subplots if num_cols * num_rows > num features
    for j in range(i + 1, len(axes)):
        fig.delaxes(axes[j])
    
    
    plt.tight_layout()
    plt.show()

def standardize_dataset(df_X, y, TEST_SIZE_PERCENTAGE, RANDOM_STATE, dict_standard_scaler, scaler_description):
    try:
        
        ############################################################################################
        # create hold-out
        ############################################################################################
        X_train, X_test, y_train, y_test = train_test_split(df_X, y, test_size=TEST_SIZE_PERCENTAGE, random_state=RANDOM_STATE)
        
        
        print(f'X_train: {X_train.shape}')
        print(f'y_train: {len(y_train)}\n')
        print(f'X_test: {X_test.shape}')
        print(f'y_test: {len(y_test)}')
        
        ############################################################################################
        # Standardize Dataset
        ############################################################################################
        SCALE_DATA = True
        
        # assign scaler
        scaler = StandardScaler()
        
        # instantiate scaler, fit, and scale X_train
        scaler = StandardScaler()
        X_train_scaled = scaler.fit_transform(X_train)
        X_test_scaled = scaler.transform(X_test)

        # store the scaler - i.e., store the description of this scaler used for outlier transformation e.g., 'before_transformation', 'iqr', 'zscore', etc
        dict_standard_scaler[scaler_description] = scaler
        
        # convert scaled numpy array to pd    
        df_X_train = pd.DataFrame(X_train_scaled, columns=X_train.columns)
        df_X_test = pd.DataFrame(X_test_scaled, columns=X_test.columns)
        
        return df_X_train, df_X_test, y_train, y_test
        
    except Exception as e:
        print('\n\n* * * EXCEPTION caught in instantiate_and_model_data mtd')
        print(f"Exception message: {e}")  # Print the exception message
        traceback.print_exc()  # Print the full stack trace

def print_df_styled_widget(df_to_print, PRECISION):
    styled_df = df_to_print.style.format(precision=PRECISION, thousands=",", decimal=".").format_index(str.lower, axis=1)
    display(HTML(styled_df.to_html()))

df = pd.read_csv('avocado_original.csv', sep=',')

df

list(df.columns)

['Unnamed: 0',
 'Date',
 'AveragePrice',
 'Total Volume',
 '4046',
 '4225',
 '4770',
 'Total Bags',
 'Small Bags',
 'Large Bags',
 'XLarge Bags',
 'type',
 'year',
 'region']

#   old name: new name
col_rename = {'Unnamed: 0': 'unnamed_0',
              'Date' : 'date',
              'AveragePrice': 'average_price',
              'Total Volume' : 'total_volume', 
              'Total Bags' : 'total_bags',
              'Small Bags' : 'small_bags',                    
              'Large Bags' : 'large_bags',
              'XLarge Bags' : 'xlarge-bags',
             }

# df = df.rename(columns=col_rename)
# or
df.rename(columns=col_rename, inplace=True)
df

try:
    df.drop('unnamed_0', axis=1, inplace=True)
except:
    pass

df

df.to_csv('./avocado.csv', sep=',', index=False)

# this can be a very useful function to print contents of dataframe when the default printing scheme doesn't display the df in an easy to read manner
def print_df(dataframe):
    with pd.option_context('display.max_rows', None, 'display.max_columns', None):
        print(tabulate(dataframe, headers='keys', tablefmt='pretty'))

#print_df(df.describe())

df.head(6)

df.tail(6)

df.sample(7, replace=False)

display(df)

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18249 entries, 0 to 18248
Data columns (total 13 columns):
 #   Column         Non-Null Count  Dtype  
---  ------         --------------  -----  
 0   date           18249 non-null  object 
 1   average_price  18249 non-null  float64
 2   total_volume   18249 non-null  float64
 3   4046           18249 non-null  float64
 4   4225           18249 non-null  float64
 5   4770           18249 non-null  float64
 6   total_bags     18249 non-null  float64
 7   small_bags     18249 non-null  float64
 8   large_bags     18249 non-null  float64
 9   xlarge-bags    18249 non-null  float64
 10  type           18249 non-null  object 
 11  year           18249 non-null  int64  
 12  region         18249 non-null  object 
dtypes: float64(9), int64(1), object(3)
memory usage: 1.8+ MB

df

df.nunique()

date               169
average_price      259
total_volume     18237
4046             17702
4225             18103
4770             12071
total_bags       18097
small_bags       17321
large_bags       15082
xlarge-bags       5588
type                 2
year                 4
region              54
dtype: int64

# df[lst_all_numerical_cols].value_counts()

df.describe().T

DEPENDENT_VAR_NAME = 'average_price'
print(f'dependent var set to --> {DEPENDENT_VAR_NAME}\n')

# all non-numeric cols
lst_non_numerical_cols = ['date', 'type', 'region']

# categorical vals
lst_categorical_cols = ['region', 'type']

# time series
lst_time_series_cols = ['date', 'year']

# all numeric cols (including ground truth if applicable)
lst_all_numerical_cols = df.columns.drop(lst_categorical_cols).drop(lst_time_series_cols)

# assign dependent var
lst_features = df[lst_all_numerical_cols].columns.drop(DEPENDENT_VAR_NAME)
print('lst_features --> ', list(lst_features), '\n')


# create df_X_features
df_X_features = df[lst_features]
print(f'dataframe of X features is named --> df_X_features:')
print_df(df_X_features.head(5))

# separate function to print the dataframe
print('\n\n')
print_df_styled_widget(df_X_features.head(5), 2)

# set dict of scalers
dict_standard_scalers = dict()
print('\ndict_standard_scalers has been initialized.')

dependent var set to --> average_price

lst_features -->  ['total_volume', '4046', '4225', '4770', 'total_bags', 'small_bags', 'large_bags', 'xlarge-bags'] 

dataframe of X features is named --> df_X_features:
+---+--------------+---------+-----------+-------+------------+------------+------------+-------------+
|   | total_volume |  4046   |   4225    | 4770  | total_bags | small_bags | large_bags | xlarge-bags |
+---+--------------+---------+-----------+-------+------------+------------+------------+-------------+
| 0 |   64236.62   | 1036.74 | 54454.85  | 48.16 |  8696.87   |  8603.62   |   93.25    |     0.0     |
| 1 |   54876.98   | 674.28  | 44638.81  | 58.33 |  9505.56   |  9408.07   |   97.49    |     0.0     |
| 2 |  118220.22   |  794.7  | 109149.67 | 130.5 |  8145.35   |  8042.21   |   103.14   |     0.0     |
| 3 |   78992.15   | 1132.0  | 71976.41  | 72.58 |  5811.16   |   5677.4   |   133.76   |     0.0     |
| 4 |   51039.6    | 941.48  | 43838.39  | 75.78 |  6183.95   |  5986.26   |   197.69   |     0.0     |
+---+--------------+---------+-----------+-------+------------+------------+------------+-------------+

dict_standard_scalers has been initialized.

df[lst_all_numerical_cols].var(axis=0, ddof=0)

average_price    1.621395e-01
total_volume     1.192632e+13
4046             1.600110e+12
4225             1.449826e+12
4770             1.154789e+10
total_bags       9.726208e+11
small_bags       5.567519e+11
large_bags       5.951613e+10
xlarge-bags      3.130214e+08
dtype: float64

analysis_start_time = datetime.datetime.now()
analysis_start_time_text = get_time()
print(f'ready for EDA timestamp: {analysis_start_time_text}')

ready for EDA timestamp: 2025-02-24-0709

df[lst_features]

df_X_features

df_describe = df_X_features.describe().T
df_describe['skew'] = df_X_features.skew()
df_describe['kurtosis'] = df_X_features.kurt()

df_describe

df_to_analyze = df[lst_all_numerical_cols]

df_to_analyze[df_to_analyze.columns[0]].hist(alpha=0.7, rwidth=0.9, edgecolor='black', density=False);
#sns.kdeplot(df_to_analyze[df_to_analyze.columns[0]], color='red', fill=False);

df_to_analyze = df[lst_all_numerical_cols]

df_to_analyze[df_to_analyze.columns[0]].hist(alpha=0.7, rwidth=0.9, edgecolor='black', density=True);
#sns.kdeplot(df_to_analyze[df_to_analyze.columns[0]], color='red', fill=False);

df_to_analyze = df[lst_all_numerical_cols]

df_to_analyze[df_to_analyze.columns[0]].hist(alpha=0.7, rwidth=0.9, edgecolor='black', density=True);
sns.kdeplot(df_to_analyze[df_to_analyze.columns[0]], color='red', fill=False);

%%time

# create 2X2 grid for subplots
#fig, axes = plt.subplots(2, 2, figsize=(10, 8), sharex=True, sharey=True)
#fig, axes = plt.subplots(2, 2, figsize=(10, 8)) # uncomment above to share same x and y scale

# set df to analyze
df_to_analyze = df_X_features


# set num rows and cols
num_cols = 3
num_rows = int(np.ceil(len(df_to_analyze.columns)/num_cols))

# create grid for subplots
fig, axes = plt.subplots(num_rows, num_cols, figsize=(15, 4*num_cols), sharex=False, sharey=False)

# flatten subplot for iteration
axes = axes.flatten()


# plot histogram in each subplot
for i, attr in enumerate(df_to_analyze.columns):
    #axes[i].hist(df_to_analyze[attr], bins=10, edgecolor='black', alpha=0.7) # can specify number of bins
    #axes[i].hist(df_to_analyze[attr], edgecolor='black', alpha=0.7) #df_to_analyze[attr], ax=axes[i], 
    
    sns.histplot(df_to_analyze[attr], bins=10, kde=False, edgecolor='black', alpha=0.7, ax=axes[i])

    
    
    axes[i].set_title(f'Histogram of {attr}')
    axes[i].set_xlabel(attr)
    axes[i].set_ylabel('Frequency')
    axes[i].tick_params(axis='both', which='both', labelsize=10)
    axes[i].grid(True, axis='y', linestyle='--')

# Hide unused subplots if num_cols * num_rows > num features
for j in range(i + 1, len(axes)):
    fig.delaxes(axes[j])


plt.tight_layout()
plt.show()

CPU times: user 1.79 s, sys: 7.8 ms, total: 1.8 s
Wall time: 4.49 s

%%time

# create 2X2 grid for subplots
#fig, axes = plt.subplots(2, 2, figsize=(10, 8), sharex=True, sharey=True)
#fig, axes = plt.subplots(2, 2, figsize=(10, 8)) # uncomment above to share same x and y scale

# set df to analyze
df_to_analyze = df_X_features


# set num rows and cols
num_cols = 3
num_rows = int(np.ceil(len(df_to_analyze.columns)/num_cols))

# create grid for subplots
fig, axes = plt.subplots(num_rows, num_cols, figsize=(15, 4*num_cols), sharex=False, sharey=False)

# flatten subplot for iteration
axes = axes.flatten()


# plot histogram in each subplot
for i, attr in enumerate(df_to_analyze.columns):
    #axes[i].hist(df_to_analyze[attr], bins=10, edgecolor='black', alpha=0.7) # can specify number of bins
    #axes[i].hist(df_to_analyze[attr], edgecolor='black', alpha=0.7) #df_to_analyze[attr], ax=axes[i], 
    
    sns.histplot(df_to_analyze[attr], bins=10, kde=True, edgecolor='black', alpha=0.7, ax=axes[i])

    
    
    axes[i].set_title(f'Histogram of {attr}')
    axes[i].set_xlabel(attr)
    axes[i].set_ylabel('Frequency')
    axes[i].tick_params(axis='both', which='both', labelsize=10)
    axes[i].grid(True, axis='y', linestyle='--')

# Hide unused subplots if num_cols * num_rows > num features
for j in range(i + 1, len(axes)):
    fig.delaxes(axes[j])


plt.tight_layout()
plt.show()

CPU times: user 2.92 s, sys: 11.9 ms, total: 2.93 s
Wall time: 9.88 s

################################################
# call fcn - density
################################################
display_hist_kde_plot(df_X_features, False, 3, "Avacado Density Plot")

%%time

################################################
# call fcn - frequency
################################################
display_hist_kde_plot(df_X_features, True, 3, "Avacado Distribution Plot - Frequencies")

CPU times: user 3.23 s, sys: 10.7 ms, total: 3.24 s
Wall time: 8.95 s

################################################
# call fcn - density
################################################
plot_kde_histplot(df[lst_all_numerical_cols], True, 3, 'Avocado Price Distibution of each Attribute')
plt.tight_layout()

################################################
# call fcn - frequency
################################################
plot_kde_histplot(df[lst_all_numerical_cols], False, 3, 'Avocado Price Distibution of each Attribute')
plt.tight_layout()

%%time

def plot_boxplot(df_to_analyze, plot_title, include_swarmplot):
    # set num rows
    num_cols = 4
    num_rows = int(np.ceil(len(df_to_analyze.columns)/num_cols))
    
    fig, axes = plt.subplots(num_rows, num_cols, figsize=(15, 5 * num_rows))  
    axes = axes.flatten()  # Flatten axes for easy iteration

    fig.suptitle(plot_title,y=0.99, size=20)
    
    for i, attr in enumerate(df_to_analyze.columns):
        sns.boxplot(y=df_to_analyze[attr], ax=axes[i], color='lightblue')

        if include_swarmplot:
            sns.swarmplot(y=attr, data=df_to_analyze)
        
        axes[i].set_title(f'Boxplot of {attr}')    
        axes[i].set_xlabel(attr)
        axes[i].tick_params(axis='both', which='both', labelsize=10)
    
    
    # Hide unused subplots (if any)
    for j in range(i + 1, len(axes)):
        fig.delaxes(axes[j])
    
    plt.tight_layout()
    plt.show()

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df[lst_all_numerical_cols]
df_to_analyze = df.select_dtypes(include=[np.number])
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

CPU times: user 1.93 s, sys: 24.4 ms, total: 1.95 s
Wall time: 3.17 s

################################################
# call fcn - plotly.go
################################################
df_to_analyze = df.select_dtypes(include=[np.number])
plot_boxplot_plotly(df_to_analyze, "Boxplot of Attributes", None)
#plot_boxplot_plotly(df_to_analyze, "Boxplot of Attributes", 'blue')

df_to_analyze = df

correlation_matrix = df_to_analyze.corr(numeric_only=True)
correlation_matrix

plt.figure(figsize=(10, 10)) 
sns.heatmap(correlation_matrix, annot=True, cmap='PuBu');

sns.scatterplot(data=df, x='total_bags', y='small_bags', hue='type', edgecolor='black');

# linear model plot --> scatter plot with a regression line
# shaded region represents the 95% confidence interval (which is directly related to standard error of measurement)
sns.lmplot(data=df, x='total_bags', y='small_bags', hue='type', scatter_kws={'edgecolor': 'k', 's': 100});

sns.violinplot(x='type', y='average_price',data=df)

<Axes: xlabel='type', ylabel='average_price'>

%%time

df_to_analyze = df

sns.pairplot(df_to_analyze, diag_kind='kde', hue="type");

CPU times: user 1min 29s, sys: 127 ms, total: 1min 29s
Wall time: 2min 27s

def plot_error_bar(df_to_analyze, show_error_bar, plot_std_error, chart_title):


    # Calculate the mean and standard deviation for each feature grouped by category
    mean_values = df_to_analyze.groupby('type').mean()
    std_values = df_to_analyze.groupby('type').std()
    std_err_values = std_values / np.sqrt(df_to_analyze.groupby('type').count())
    
    # determine which bar to plot
    lst_error_bar = std_values
    
    if plot_std_error:
        lst_error_bar = std_err_values
    
    # Set up the plot
    plt.figure(figsize=(12, 8))
    
    # Plotting the mean and filling the standard deviation area for each category
    for i, category in enumerate(mean_values.index):
        if plot_error_bar:
            plt.errorbar(mean_values.columns, mean_values.iloc[i], yerr=lst_error_bar.iloc[i], 
                     label=f'{category}', marker='o', capsize=5)
        else:
            plt.plot(mean_values.columns, mean_values.iloc[i], marker='o', label=f'{category}')
        
        plt.fill_between(mean_values.columns,
                         mean_values.iloc[i] - lst_error_bar.iloc[i],
                         mean_values.iloc[i] + lst_error_bar.iloc[i],
                         alpha=0.2)  # Alpha controls the transparency of the fill
    
    # Customizing the plot
    plt.xlabel('Features')
    plt.ylabel('Value')
    plt.title(chart_title)
    plt.legend(title='Avacadoes')
    plt.xticks(rotation=45)  # Rotate x-axis labels by 45 degrees
    plt.grid(True, axis='y', linestyle='--')
    plt.show()

################################################
# call fcn - error bar
################################################
df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze['type'] = df['type']  # Add back the 'type' column
show_error_bar = True
plot_std_error = False

plot_error_bar(df_to_analyze, show_error_bar, plot_std_error, 'Mean and Standard Error for Each Type')

################################################
# call fcn - std dev
################################################
df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze['type'] = df['type']  # Add back the 'type' column
show_error_bar = True
plot_std_error = True

plot_error_bar(df_to_analyze, show_error_bar, plot_std_error, 'Mean and Standard Deviation for Each Type')

# notify
print('DEPENDENT_VAR_NAME:' + '\n' + ('='*40) + '\n' + DEPENDENT_VAR_NAME)

# assign features
print('\nfeatures\n' + ('='*40))
print(list(lst_features))

# assign y
y = df[DEPENDENT_VAR_NAME]

# notify
print('\nDependent variable y:' + '\n' + ('='*40))
print(f'y is of type {type(y)}    num_rows: {len(y)}')

# assign X
X = df[lst_features]

# notify
print('\nIndependent variables X:' + '\n' + ('='*40))
print(f'X is of type {type(X)}  num_rows: {len(X)}   num_cols: {len(X.columns)}')

DEPENDENT_VAR_NAME:
========================================
average_price

features
========================================
['total_volume', '4046', '4225', '4770', 'total_bags', 'small_bags', 'large_bags', 'xlarge-bags']

Dependent variable y:
========================================
y is of type <class 'pandas.core.series.Series'>    num_rows: 18249

Independent variables X:
========================================
X is of type <class 'pandas.core.frame.DataFrame'>  num_rows: 18249   num_cols: 8

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=42)

print(f'X_train: {X_train.shape}')
print(f'y_train: {len(y_train)}\n')

print(f'X_test: {X_test.shape}')
print(f'y_test: {len(y_test)}')

X_train: (14599, 8)
y_train: 14599

X_test: (3650, 8)
y_test: 3650

SCALE_DATA = True

# assign scaler
scaler = StandardScaler()

# instantiate scaler, fit, and scale X_train
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

# store scaler - indicate this is before outlier transformations
dict_standard_scalers['before_transformation'] = scaler

# convert scaled numpy array to pd    
df_X_train = pd.DataFrame(X_train_scaled, columns=X_train.columns)
df_X_test = pd.DataFrame(X_test_scaled, columns=X_test.columns)

analysis_start_time = datetime.datetime.now()
analysis_start_time_text = get_time()
print(f'ready to train models timestamp: {analysis_start_time_text}')

ready to train models timestamp: 2025-02-24-0714

fig, axes = plt.subplots(1, 2, figsize=(12, 5))

# Before scaling
X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")

# After scaling
df_X_train.plot(kind='kde', ax=axes[1], title="After Scaling")

plt.tight_layout()
plt.show()

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df[lst_all_numerical_cols]
df_to_analyze = df_X_train
plot_boxplot(df_to_analyze, "Boxplot of Standardized Features", False)

df_to_analyze = df_X_train
df_to_analyze[df_to_analyze.columns].hist(figsize=(20,16));

df_to_analyze = df_X_train
plot_kde_histplot(df_to_analyze, False, 3, 'Standardized Avocado Price Distibution of each Feature')

display_hist_kde_plot(df_X_features, True, 3, "Standardized Avacado Distribution Plot")

MAPE = np.mean(np.abs((y_true - y_pred) / np.maximum(np.abs(y_true), 1e-8))) * 100  # 1e-8 prevents divide by zero

SMAPE = 100 * np.mean(2 * np.abs(y_true - y_pred) / np.maximum(np.abs(y_true) + np.abs(y_pred), 1e-8))

analysis_start_time_text = "not started"
analysis_end_time_text = "not started"

analysis_start_time = datetime.datetime.now()
analysis_start_time_text = get_time()

#############################################################################################
# instantiate model
#############################################################################################
model_lr = LinearRegression()

#############################################################################################
# train model
#############################################################################################
model_lr.fit(X_train_scaled, y_train)

#############################################################################################
# inference: use trained model to predict our test set
#############################################################################################
y_pred = model_lr.predict(X_test_scaled)

#############################################################################################
# use cross validation for more accurate evaluation
#############################################################################################
cv_lr = cross_val_score(estimator = model_lr, X = X_train_scaled, y = y_train, cv = 10)

#############################################################################################
# evaluate performance of model
#############################################################################################


# Standard regression metrics
r2 = r2_score(y_test, y_pred)  
MSE = mean_squared_error(y_test, y_pred)  
RMSE = np.sqrt(MSE)  
MAE = mean_absolute_error(y_test, y_pred)  
MedAE = median_absolute_error(y_test, y_pred)
MAPE = 100 * np.mean(np.abs((y_test - y_pred) / y_test))
RAND = rand_score(y_test, y_pred)
RMSLE = np.sqrt(mean_squared_error(np.log1p(y_test), np.log1p(y_pred)))

# Adjusted R² - FIX
n = len(y_test)  
p = X_test.shape[1]  
adjusted_r2 = 1 - (1 - r2) * (n - 1) / (n - p - 1)  

# cv r2
cv_r2_lr = cv_lr.mean()

# MASE
m = 1  # seasonal period (change as needed)
naive_forecast = y_test[:-m]  # Shifted actual values
MASE = mean_absolute_error(y_test[m:], y_pred[m:]) / mean_absolute_error(y_test[m:], naive_forecast)

#SMAPE
SMAPE = 100 * np.mean(2 * np.abs(y_test - y_pred) / (np.abs(y_test) + np.abs(y_pred)))  

# execution time
analysis_end_time = datetime.datetime.now()
analysis_end_time_text = get_time()

analysis_duration = analysis_end_time - analysis_start_time
analysis_duration_text = calculate_time_duration(analysis_duration)  

#############################################################################################
# notify
#############################################################################################
print(f"R²: {r2:.4f}")
print(f"Adjusted R²: {adjusted_r2:.4f}")
print(f"CV R²: {cv_r2_lr:.4f}")

print(f"RMSE: {RMSE:.4f}")
print(f"RMSLE: {RMSLE:.4f}")

print(f"MAE: {MAE:.4f}")
print(f"MAPE: {MAPE:.4f}%")

print(f"MSE: {MSE:.4f}")
print(f"MASE: {MASE:.4f}")
print(f"SMAPE: {SMAPE:.4f}%")
print(f"MedAE: {MedAE:.4f}")

print(f"RAND: {RAND:.4f}")



print("Start Time:\t\t" + analysis_start_time_text)
print("End Time:\t\t" + analysis_end_time_text)
print("Analysis Duration:\t" + analysis_duration_text)

R²: 0.0435
Adjusted R²: 0.0414
CV R²: 0.0517
RMSE: 0.3920
RMSLE: 0.1623
MAE: 0.3147
MAPE: 24.6377%
MSE: 0.1537
MASE: 0.6788
SMAPE: 22.8480%
MedAE: 0.2767
RAND: 0.9927
Start Time:		2025-02-24-0715
End Time:		2025-02-24-0715
Analysis Duration:	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)

mdlName = 'linear_regression'
dict_coefs, plt = display_feature_importance_chart(model_lr, mdlName, list(X_train.columns), 10, 'feature_importance_' + str(mdlName), X_train_scaled, y_test, True)

dict_coefs

coefficients:
====================
total_bags	33248.058758235056
small_bags	24999.24868347553
large_bags	8219.863487202136
xlarge-bags	602.5671941924443
total_volume	543.7725435056285
4046	198.68280392629217
4225	190.85499709288683
4770	17.24034704200814

{'total_bags': '33248.058758235056',
 'small_bags': '24999.24868347553',
 'large_bags': '8219.863487202136',
 'xlarge-bags': '602.5671941924443',
 'total_volume': '543.7725435056285',
 '4046': '198.68280392629217',
 '4225': '190.85499709288683',
 '4770': '17.24034704200814'}

df_to_analyze = df_X_features

# correlation
correlation_matrix = df_to_analyze.corr(numeric_only=True)

# calculate magnitude of correlations
correlation_matrix['correlation_magnitude'] = correlation_matrix.abs().sum(axis=1)

# sort by magnitude
correlation_matrix = correlation_matrix.sort_values(by='correlation_magnitude', ascending=False)

correlation_matrix

plt.figure(figsize=(10, 10)) 
sns.heatmap(correlation_matrix.drop('correlation_magnitude', axis=1), annot=True, cmap='PuBu');

def pipeline_model(model_name, model, X_TRAIN, X_TEST, Y_TRAIN, Y_TEST):
    try:
        dict_evaluation = dict()
        ########################################
        # set time duration params
        ########################################
        analysis_start_time_text = "not started"
        analysis_end_time_text = "not started"
        
        analysis_start_time = datetime.datetime.now()
        analysis_start_time_text = get_time()
        
        ########################################
        # instantiate model
        ########################################
        # model has been instantiated
        print('\n')
        print('#'*70)
        print(f'# training model --> {model_name}')
        print('#'*70)
        
        try:
            print(f'hyperparameters --> {model.get_params()}\n')
        except:
            pass
        
        ########################################
        # train model
        ########################################
        model.fit(X_TRAIN, Y_TRAIN)
        
        ########################################
        # inference: use trained model to predict our test set
        ########################################
        y_pred = model.predict(X_TEST)
        
        ########################################
        # use cross validation for more accurate evaluation
        ########################################
        cv_model = cross_val_score(estimator=model, X=X_TRAIN, y=Y_TRAIN, cv=10)
        
        ########################################
        # evaluate performance of model
        ########################################        
        
        # General Purpose Metrics
        r2 = r2_score(Y_TEST, y_pred)
        MSE = mean_squared_error(Y_TEST, y_pred)
        RMSE = np.sqrt(MSE)
        MAE = mean_absolute_error(Y_TEST, y_pred)
        MedAE = median_absolute_error(Y_TEST, y_pred)
        EVS = explained_variance_score(Y_TEST, y_pred)
    
        # Adjusted R²
        n = len(Y_TEST)
        p = X_TEST.shape[1]
        adjusted_r2 = 1 - (1 - r2) * (n - 1) / (n - p - 1) if n > p + 1 else np.nan
    
        # CV R²
        cv_r2 = cv_model.mean()
    
        # Metrics for Specific Situations
        MAPE = np.mean(np.abs((Y_TEST - y_pred) / np.maximum(np.abs(Y_TEST), 1e-8)))  # Avoid divide-by-zero
    
        # convert to not allow negative vals
        Y_TEST_non_negative = np.maximum(Y_TEST, 0) # Clip negative values to 0
        y_pred_non_negative = np.maximum(y_pred, 0) # Clip negative predictions to 0
    
        RMSLE = np.sqrt(mean_squared_error(np.log1p(Y_TEST_non_negative), np.log1p(y_pred_non_negative)))
        MSLE = mean_squared_log_error(Y_TEST_non_negative, y_pred_non_negative)
    
        m = 1  # Seasonality (m=1 for non-seasonal data)
        naive_forecast = np.roll(Y_TEST, shift=m)
        MASE = mean_absolute_error(Y_TEST[m:], y_pred[m:]) / mean_absolute_error(Y_TEST[m:], naive_forecast[m:])
    
        SMAPE = np.mean(2 * np.abs(Y_TEST - y_pred) / (np.maximum(np.abs(Y_TEST) + np.abs(y_pred), 1e-8)))
    
        CVRMSE = (RMSE / np.mean(Y_TEST)) if np.mean(Y_TEST) != 0 else np.nan #Handle potential 0 division
    
        # Diagnostic Metrics
        MBD = np.mean(y_pred - Y_TEST) #Corrected calculation for MBD
        r2_residuals = r2_score(Y_TEST, Y_TEST - y_pred)  # R² of Residuals

        
        
        DW = durbin_watson(Y_TEST - y_pred)
    
        # Robust Metric
        huber = HuberRegressor(epsilon=1.35)
        huber.fit(X_TRAIN, Y_TRAIN)
        y_pred_huber = huber.predict(X_TEST)
        HUBER_LOSS = np.mean(np.abs(Y_TEST - y_pred_huber))  # MAE for Huber predictions
    
    
        
        # execution time
        analysis_end_time = datetime.datetime.now()
        analysis_end_time_text = get_time()
        
        analysis_duration = analysis_end_time - analysis_start_time
        analysis_duration_text = calculate_time_duration(analysis_duration)  
        
        ########################################
        # store data
        ########################################
        dict_evaluation['model_name'] = model_name
        dict_evaluation['r2'] = f'{r2:.4f}'
        dict_evaluation['adjusted_r2'] = f'{adjusted_r2:.4f}'
        dict_evaluation['cross_validated_r2'] = f'{cv_r2:.4f}'
        dict_evaluation['explained_variance_score'] = f'{EVS:.4f}'
    
        dict_evaluation['rmse'] = f'{RMSE:.4f}'        
        dict_evaluation['rmsle'] = f'{RMSLE:.4f}'
        dict_evaluation['mae'] = f'{MAE:.4f}'
        dict_evaluation['mape'] = f'{MAPE:.4f}'
        dict_evaluation['mse'] = f'{MSE:.4f}'
        dict_evaluation['msle'] = f'{MSLE:.4f}'
        dict_evaluation['mase'] = f'{MASE:.4f}'
        dict_evaluation['smape'] = f'{SMAPE:.4f}'
        dict_evaluation['med_ae'] = f'{MedAE:.4f}'
    
        dict_evaluation['huber_loss'] = HUBER_LOSS
        dict_evaluation['mbd'] = f'{MBD:.4f}'
        dict_evaluation['cv_rmse'] = f'{CVRMSE:.4f}'
        dict_evaluation['r2_residuals'] = f'{r2_residuals:.4f}'
        dict_evaluation['durbin_watson_stat'] = f'{DW:.4f}'
        
        dict_evaluation['start_time'] = analysis_start_time_text
        dict_evaluation['end_time'] = analysis_end_time_text
        dict_evaluation['analysis_duration_text'] = analysis_duration_text
        dict_evaluation['analysis_duration'] = str(analysis_duration)

        ########################################
        # notify
        ########################################
        print(f'training model {model_name} complete! \tR2: {r2:.4f}\tanalysis_duration --> {analysis_duration}\n')
    
        ########################################
        # return 
        ########################################
        return dict_evaluation
        
    except Exception as e:
        print('\n\n* * * EXCEPTION WHILE EXECUTING MODEL', model_name, '\n\n\n')
        print(f"Exception message: {e}")  # Print the exception message
        traceback.print_exc()  # Print the full stack trace

model_lr = LinearRegression()
dict_evaluation_lr = pipeline_model('linear_regression', model_lr, X_train_scaled, X_test_scaled, y_train, y_test)

dict_evaluation_lr


######################################################################
# training model --> linear_regression
######################################################################
hyperparameters --> {'copy_X': True, 'fit_intercept': True, 'n_jobs': None, 'positive': False}

training model linear_regression complete! 	R2: 0.0435	analysis_duration --> 0:00:00.285648

{'model_name': 'linear_regression',
 'r2': '0.0435',
 'adjusted_r2': '0.0414',
 'cross_validated_r2': '0.0517',
 'explained_variance_score': '0.0436',
 'rmse': '0.3920',
 'rmsle': '0.1618',
 'mae': '0.3147',
 'mape': '0.2464',
 'mse': '0.1537',
 'msle': '0.0262',
 'mase': '0.6788',
 'smape': '0.2285',
 'med_ae': '0.2767',
 'huber_loss': 0.31392678854682393,
 'mbd': '-0.0053',
 'cv_rmse': '0.2779',
 'r2_residuals': '-11.3516',
 'durbin_watson_stat': '2.0590',
 'start_time': '2025-02-24-0715',
 'end_time': '2025-02-24-0715',
 'analysis_duration_text': '0 day(s), 0 hour(s), 0 minute(s), 0 second(s)',
 'analysis_duration': '0:00:00.285648'}

mdlName = 'linear_regression'
dict_coeff_lr, plt = display_feature_importance_chart(model_lr, mdlName, list(X_train.columns), 10, 'feature_importance_' + str(mdlName), X_test_scaled, y_test, True)

coefficients:
====================
total_bags	33248.058758235056
small_bags	24999.24868347553
large_bags	8219.863487202136
xlarge-bags	602.5671941924443
total_volume	543.7725435056285
4046	198.68280392629217
4225	190.85499709288683
4770	17.24034704200814

def instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test):
    try:
        lst_evaluation_metrics = ['model_name', 
                                  'r2', 
                                  'adjusted_r2', 
                                  'cross_validated_r2', 
                                  'explained_variance_score', 'rmse', 
                                  'rmsle', 
                                  'mae', 
                                  'mape', 
                                  'mse', 
                                  'msle', 
                                  'mase', 
                                  'smape', 
                                  'med_ae', 
                                  'huber_loss', 
                                  'mbd', 
                                  'cv_rmse', 
                                  'r2_residuals', 
                                  'durbin_watson_stat', 
                                  'start_time', 
                                  'end_time', 
                                  'analysis_duration_text', 
                                  'analysis_duration'
                                 ]
        
        RANDOM_STATE = 42
        
        # create dataframe to contain evaluation metrics
        df_model_evaluation_metrics = pd.DataFrame(columns=lst_evaluation_metrics)
        
        # notify['model_name', 'r2', 'adjusted_r2', 'cross_validated_r2', 'explained_variance_score', 'rmse', 'rmsle', 'mae', 'mape', 'mse', 'msle', 'mase', 'smape', 'med_ae', 'huber_loss', 'mbd', 'cv_rmse', 'r2_residuals', 'durbin_watson_stat', 'start_time', 'end_time', 'analysis_duration_text', 'analysis_duration']
        print('instantiating models...')
        
        ada_boost_reg = AdaBoostRegressor(random_state=RANDOM_STATE)
        ard_regression = ARDRegression()
        bayesian_ridge = BayesianRidge()
        
        catboost_reg = CatBoostRegressor(random_state=RANDOM_STATE, verbose=0)
        decision_tree_regressor = DecisionTreeRegressor(random_state=RANDOM_STATE)
        decision_tree_regressor_pruned = DecisionTreeRegressor(random_state=RANDOM_STATE, max_depth=5)
        elasticnet_cv = ElasticNetCV(random_state=RANDOM_STATE)
        elasticnet_cv2 = ElasticNetCV(random_state=RANDOM_STATE, alphas=[0.5],cv = 11,n_jobs = -1)
        elasticnet_reg = ElasticNet(random_state=RANDOM_STATE)
        elasticnet_reg2 = ElasticNet(random_state=RANDOM_STATE, alpha=0.1, l1_ratio=0.5)
        extra_trees_reg = ExtraTreesRegressor(random_state=RANDOM_STATE)
        
        gaussian_process_reg = GaussianProcessRegressor(random_state=RANDOM_STATE)
        gbr_reg = GradientBoostingRegressor(random_state=RANDOM_STATE, verbose=False)
        gbr2 = GradientBoostingRegressor(random_state=RANDOM_STATE, n_estimators=300,min_weight_fraction_leaf=0.5,max_depth=2,learning_rate=0.01)
        gbr3 = GradientBoostingRegressor(random_state=RANDOM_STATE, max_depth=10, n_estimators=200, learning_rate=1.0)
        hist_gb_reg = HistGradientBoostingRegressor(random_state=RANDOM_STATE)
        huber_reg = HuberRegressor() 
        
        knn_reg_1 = KNeighborsRegressor()
        knn_reg_2 = KNeighborsRegressor(n_neighbors=5)
        
        lasso_cv = LassoCV(random_state=RANDOM_STATE, alphas=[1],cv = 11,n_jobs = -1) # cv - number of cross validations to be checked, n_jobs = -1 - use all the parallel processing power of the system
        lasso_reg2 = Lasso(random_state=RANDOM_STATE, alpha=0.1) #Parameterized Regressors
        lasso_regressor = Lasso(random_state=RANDOM_STATE)
        
        lgbm_reg = LGBMRegressor(random_state=RANDOM_STATE)
        linear_regressor = LinearRegression()
        
        mlp_reg_1 = MLPRegressor(random_state=RANDOM_STATE, max_iter=100)
        mlp_reg_2 = MLPRegressor(random_state=RANDOM_STATE, hidden_layer_sizes=(200,), max_iter=400)
        
        ransac_reg = RANSACRegressor(random_state=RANDOM_STATE)
        
        rf_reg = RandomForestRegressor(random_state=RANDOM_STATE, verbose=0, n_jobs=-1)
        rf_reg2 = RandomForestRegressor(random_state=RANDOM_STATE, n_estimators=100,min_samples_leaf=5) # n_estimators - The number of trees in the forest, min_samples_leaf - The minimum number of samples required to be at a leaf node.
        
        ridge_cv = RidgeCV(alphas=[2],cv = 11)
        ridge_reg = Ridge(random_state=RANDOM_STATE)
        ridge_reg2 = Ridge(random_state=RANDOM_STATE, alpha=1, solver="cholesky")
        
        svm_poly_reg = SVR(kernel="poly", degree=2, C=100, epsilon=0.1)
        svm_reg = SVR()
        svm_reg_lin = LinearSVR()
        svm_reg_lin2 = LinearSVR(epsilon=1.5)
        svm_reg_poly = SVR(kernel='poly') # non-linear kernel
        svm_reg_rbf = SVR(kernel='rbf') # non-linear kernel
        svm_reg_sigmoid = SVR(kernel='sigmoid') # non-linear kernel
        
        theil_sen_reg = TheilSenRegressor(max_iter=500, max_subpopulation=10000, n_jobs=-1)
        
        xgb_reg = XGBRegressor(random_state=RANDOM_STATE)
        xgb2 = XGBRegressor(random_state=RANDOM_STATE, eval_metric='logloss')
        
        
        param_grid = [   {'n_estimators': [3, 10, 30], 'max_features': [2, 4, 6, 8]},
                         {'bootstrap': [False], 'n_estimators': [3, 10], 'max_features': [2, 3, 4]},
                     ]
        grid_search = GridSearchCV(rf_reg2, param_grid, cv=5, scoring='neg_mean_squared_error',return_train_score=True)
        
        
        
        #Tuple analysis Models
        tpl_models = [
                      ('ada_boost_regressor', ada_boost_reg),              
                      ('ard_regressor', ard_regression), 
                      ('bayesian_ridge_regressor', bayesian_ridge),
                      ('decision_tree_regressor', decision_tree_regressor), 
                      ('decision_tree_pruned_regressor', decision_tree_regressor_pruned),                      
                      ('elasticnet_regressor', elasticnet_reg),
                      ('elasticnet_regressor_2', elasticnet_reg2),
                      ('elasticnet_regressor_cv', elasticnet_cv),
                      ('elastic_regressor_cv_2', elasticnet_cv2),      
                      ('extra_trees_regressor', extra_trees_reg),
                      ('gradient_boosting_regressor_1', gbr_reg),
                      ('gradient_boosting_regressor_2', gbr2),
                      ('hist_gradient_boosting_regressor', hist_gb_reg),                            
                      ('huber_regressor', huber_reg),
                      ('knn_regressor_1', knn_reg_1),   
                      ('knn_regressor_2', knn_reg_2), 
                      ('linear_regressor', linear_regressor),
                      ('lasso_regressor', lasso_regressor),
                      ('lasso_regressor_cv', lasso_cv),
                      ('lasso_regressor_2', lasso_reg2),
                      ('lgbm_regressor', lgbm_reg),
                      ('mlp_regressor_1', mlp_reg_1),
                      ('ransac_regressor', ransac_reg),                                         
                      ('ridge_regressor', ridge_reg),
                      ('ridge_regressor_cv', ridge_cv),
                      ('ridge_regressor_2', ridge_reg2),
                      ('svm_regressor_linear_1', svm_reg_lin),
                      ('svm_regressor_linear_2', svm_reg_lin2),     
                      ('theil_sen_regressor', theil_sen_reg),
                      ('xgb_regressor_1', xgb_reg),  
                      ('xgb_regressor_2', xgb2),
                  
                      
                      # longer running models                                                
                      ('random_forest_regressor', rf_reg),
                      ('random_forest_regressor_2', rf_reg2),   
                      ('mlp_regressor_2', mlp_reg_2),
                      ('catboost_regressor', catboost_reg),     
        
        
                        # NOT TESTED - execution time too long for right now
             #         ('gaussian_process_regressor', gaussian_process_reg),
             #         ('gradient_boosting_regressor_3', gbr3),
             #         ('svm_regressor', svm_reg),              
             #         ('svm_rbf_regressor', svm_reg_rbf),
             #         ('svm_poly_regressor', svm_reg_poly),
             #         ('svm_sigmoid_regressor', svm_reg_sigmoid),
             #         ('grid_search_regressor', grid_search),
            
                     ]
        
        print('model instantiation complete!')
        
        
        lst_error_model = []
        dict_completed_models = dict()
        
        # train the models
        #for model_name, model in tpl_models:
        for model_name, model in tqdm(tpl_models, desc=f"Processing Model", unit="model"):
            try:
                
                # train and evaluate model
                dict_evaluation = pipeline_model(model_name, model, X_train_scaled, X_test_scaled, y_train, y_test)
            
                # store instance
                if dict_evaluation is not None:
                    df_model_evaluation_metrics.loc[len(df_model_evaluation_metrics)] = dict_evaluation
                    dict_completed_models[model_name] = model
                else:
                    lst_error_model.append(model_name)    
                        
            except Exception as e:        
                lst_error_model.append(model_name) 
                print('\n\n* * * EXCEPTION WHILE EXECUTING MODEL', model_name)
                print(f"Exception message: {e}")  # Print the exception message
                traceback.print_exc()  # Print the full stack trace
                
                continue
        
        # notify training complte
        print('\n\n')
        print('*'*90)
        print('* model training COMPLETE!')
        print('*'*90)
          
              
        # notify training error
        if len(lst_error_model) > 0:
            print('\n\n')
            print('*'*90)
            print('* * * ERROR MODELS:')
            print('*'*90)
        
            for model_name in lst_error_model:
                print(f'{model_name}')

        return df_model_evaluation_metrics, dict_completed_models, lst_error_model
        
        print('\n\n')
    except Exception as e:
        print('\n\n* * * EXCEPTION caught in instantiate_and_model_data mtd')
        print(f"Exception message: {e}")  # Print the exception message
        traceback.print_exc()  # Print the full stack trace

%%time 
df_model_evaluation_metrics, dict_completed_models, lst_error_model = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:08<04:40,  8.26s/model]

training model ada_boost_regressor complete! 	R2: 0.3543	analysis_duration --> 0:00:08.258497


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   6%|██▍                                        | 2/35 [00:08<02:01,  3.67s/model]

training model ard_regressor complete! 	R2: 0.0430	analysis_duration --> 0:00:00.456839


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:08<01:07,  2.10s/model]

training model bayesian_ridge_regressor complete! 	R2: 0.0433	analysis_duration --> 0:00:00.210436


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

Processing Model:  11%|████▉                                      | 4/35 [00:11<01:07,  2.19s/model]

# sort models by r2
df_model_evaluation_metrics = df_model_evaluation_metrics.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics = df_model_evaluation_metrics.reset_index(drop=True, inplace=False)

#save df
df_model_evaluation_metrics.to_csv('./model_performance_before_outlier_removal.csv', index=False)

# view df
df_model_evaluation_metrics

# plot in ascending order
df_model_evaluation_metrics_sorted = df_model_evaluation_metrics.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted = df_model_evaluation_metrics_sorted.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted = dict()

lst_sorted_models = list(df_model_evaluation_metrics_sorted['model_name'])[::-1]

# plot
plt.figure(figsize=(20, 10))  # Adjust figure size if needed (width, height)
plt.barh(df_model_evaluation_metrics_sorted['model_name'], df_model_evaluation_metrics_sorted['r2'], edgecolor='black') # barh for horizontal

plt.xlabel("R-squared")  # Note the x and y labels are switched
plt.ylabel("Model Name")
plt.title("Model Performance Based on R-squared")

plt.xticks(rotation=45)
plt.grid(axis='x', linestyle='--', alpha=0.7)
#plt.xlim(0, None)

plt.tight_layout()
plt.show()

fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted['model_name'],  
    x=df_model_evaluation_metrics_sorted['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()

%%time

for model_name in tqdm(lst_sorted_models, desc=f"Processing Model Feature Importancen", unit="model"):
    try:
        model = dict_completed_models[model_name]

        r2 = df_model_evaluation_metrics[df_model_evaluation_metrics['model_name'] == model_name].r2.iloc[0]
        
        try:
            
            print(f'\nmodel [{model_name}] R2 --> {r2}')
            print('='*70, '\n')
        except:
            print(f"Exception message: {e}")  # Print the exception message
            traceback.print_exc()  # Print the full stack trace
        
        # get feature importance
        lst_coefficients, plt = display_feature_importance_chart(model, model_name, list(X_train.columns), 10, 'feature_importance_' + str(model_name), X_test_scaled, y_test, True)
    except Exception as e:
        print('Exception caught while processing model:', model_name)
        print(f"Exception message: {e}")  # Print the exception message
        traceback.print_exc()  # Print the full stack trace
        continue

print('\n\n')

Processing Model Feature Importancen:   0%|                               | 0/35 [00:00<?, ?model/s]

model [extra_trees_regressor] R2 --> 0.7657
====================================================================== 

coefficients:
====================
4046	0.2044756525504432
total_volume	0.1538843229161739
4225	0.14097137924697012
large_bags	0.13452144138631242
total_bags	0.1285978523252027
small_bags	0.11051697662718887
4770	0.09095538353290719
xlarge-bags	0.03607699141480161

Processing Model Feature Importancen:   3%|▋                      | 1/35 [00:01<00:39,  1.17s/model]

model [random_forest_regressor] R2 --> 0.7634
====================================================================== 

coefficients:
====================
total_volume	0.3839449501662739
4046	0.16137563461599813
4225	0.12285801829382291
large_bags	0.11081643801894067
small_bags	0.07349521127498083
4770	0.07021993324258416
total_bags	0.049181414573259925
xlarge-bags	0.028108399814139446

Processing Model Feature Importancen:   6%|█▎                     | 2/35 [00:02<00:37,  1.14s/model]

model [random_forest_regressor_2] R2 --> 0.7441
====================================================================== 

coefficients:
====================
total_volume	0.41702970228251524
4046	0.15979554757490042
4225	0.115591731171966
large_bags	0.10645224442885194
small_bags	0.06664889503045228
4770	0.06656442954624363
total_bags	0.04129471644664914
xlarge-bags	0.026622733518421274

Processing Model Feature Importancen:   9%|█▉                     | 3/35 [00:02<00:30,  1.07model/s]

model [catboost_regressor] R2 --> 0.7264
====================================================================== 

coefficients:
====================
4046	21.14370965027229
4225	18.91334545807429
large_bags	14.642110949181259
total_volume	12.079962507875798
4770	11.2786945951956
small_bags	11.112353483462769
total_bags	6.241216351876749
xlarge-bags	4.588607004061248

Processing Model Feature Importancen:  11%|██▋                    | 4/35 [00:03<00:26,  1.17model/s]

model [xgb_regressor_2] R2 --> 0.7201
====================================================================== 

coefficients:
====================
total_volume	0.36144048
4046	0.1252322
large_bags	0.12117669
4225	0.093668796
4770	0.08301563
xlarge-bags	0.07807932
small_bags	0.07747827
total_bags	0.059908584

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df[lst_all_numerical_cols]
#df_to_analyze = df.select_dtypes(include=[np.number])
df_X_features = df[lst_features]
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

# create df_X_features
df_X_features = df[lst_features]
print(f'dataframe of X features is named --> df_X_features:')
#print_df(df_X_features.head(5))
df_X_features

dataframe of X features is named --> df_X_features:

df_describe = df_X_features.describe().T
df_describe['skew'] = df_X_features.skew()
df_describe['kurtosis'] = df_X_features.kurt()

df_describe

%%time

fig, axes = plt.subplots(1, 2, figsize=(12, 5))

# Before scaling
X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")

# After scaling
df_X_train.plot(kind='kde', ax=axes[1], title="After Scaling")

plt.tight_layout()
plt.show()

print('\n')


CPU times: user 8.97 s, sys: 432 μs, total: 8.97 s
Wall time: 11.2 s

df_to_analyze = df_X_features
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

def detect_and_remove_outliers_iqr(df_to_analyze, lst_features):
    """
    Detects and removes outliers from specified columns in a DataFrame using the IQR method.

    Args:
        df_to_analyze (pd.DataFrame): The input DataFrame.
        lst_features (list): A list of column names to analyze for outliers.

    Returns:
        pd.DataFrame: The DataFrame with outliers removed from all specified columns.
        dict: A dictionary where keys are column names and values are lists of outlier indices for each column.
    """

    all_outlier_indices = {feature: [] for feature in lst_features}  # Initialize all_outlier_indices
    df_cleaned = df_to_analyze.copy()

    count_total_outliers_removed = 0

    for feature_name in lst_features:
        try:
            df_cleaned[feature_name] = pd.to_numeric(df_cleaned[feature_name], errors='coerce')
            df_cleaned.dropna(subset=[feature_name], inplace=True)

            data = df_cleaned[feature_name].values

            if len(data) < 3:
                continue #Skip to the next feature

            q1 = np.percentile(data, 25)
            q3 = np.percentile(data, 75)
            iqr = q3 - q1

            lower_bound = q1 - 1.5 * iqr
            upper_bound = q3 + 1.5 * iqr

            outlier_mask = (data < lower_bound) | (data > upper_bound) #Create a boolean mask
            
            # accumulate list of outlier indicies for this feature
            all_outlier_indices[feature_name] = df_cleaned.index[outlier_mask].tolist() #Use the mask with the original index

            count_outliers_removed_within_this_feature = len(all_outlier_indices[feature_name])
            print(f'Count outliers to remove from feature: [{feature_name}] --> {count_outliers_removed_within_this_feature}')

            

        except Exception as e:
            print(f"Error processing column '{feature_name}': {e}")
            continue  # Continue to the next feature

    #Remove all outlier rows at once, using the collected indices.
    rows_to_remove = []
    
    # accumulate all indicies for all features first
    for indices in all_outlier_indices.values():
        rows_to_remove.extend(indices)

    # Remove duplicates and sort
    rows_to_remove = sorted(list(set(rows_to_remove))) 

    count_total_outliers_removed = len(rows_to_remove)

    df_cleaned = df_to_analyze.drop(rows_to_remove, errors='ignore') #Drop from original dataframe using the original indices

    return df_cleaned, all_outlier_indices, count_total_outliers_removed

# NOTE: AI assisted code from Gemini

# detecting outliers
df_cleaned, outlier_indices_dict, count_total_outliers_removed = detect_and_remove_outliers_iqr(df, lst_features)
y_cleaned = df_cleaned[DEPENDENT_VAR_NAME]
print(f'\ntotal number of outlier instances removed: --> {count_total_outliers_removed}')

# set cleaned df
df_cleaned_features_IQR = df_cleaned[lst_features]

Count outliers to remove from feature: [total_volume] --> 2297
Count outliers to remove from feature: [4046] --> 2518
Count outliers to remove from feature: [4225] --> 2519
Count outliers to remove from feature: [4770] --> 2918
Count outliers to remove from feature: [total_bags] --> 2533
Count outliers to remove from feature: [small_bags] --> 2407
Count outliers to remove from feature: [large_bags] --> 2707
Count outliers to remove from feature: [xlarge-bags] --> 4014

total number of outlier instances removed: --> 6502

###############################################
# print original skewness and kurtosis
###############################################
SKEWNESS = skew(df_X_features)
KURTOSIS = kurtosis(df_X_features)
shape_metric = list(np.sqrt(np.power(SKEWNESS, 2) + np.power(KURTOSIS, 2)))
sum_shape_metric = np.sum(shape_metric)

#print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} \t shape_metric: {shape_metric}')
print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} ')

###############################################
# print transformed skewness and kurtosis
###############################################
SKEWNESS_TRANSFORMED = skew(df_cleaned_features_IQR)
KURTOSIS_TRANSFORMED = kurtosis(df_cleaned_features_IQR)
shape_metric_TRANSFORMED = list(np.sqrt(np.power(SKEWNESS_TRANSFORMED, 2) + np.power(KURTOSIS_TRANSFORMED, 2)))
sum_shape_metric_TRANSFORMED = np.sum(shape_metric_TRANSFORMED)

#print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED} \t shape_metric: {shape_metric_TRANSFORMED}')
print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED}')

################################################
# call fcn - sns hist plots
################################################
df_to_analyze = df_cleaned_features_IQR
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze = df_cleaned_features_IQR
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

ORIGINAL    sum_shape_metric: 977.3584239619304 
TRANSFORMED sum_shape_metric: 104.91148475412267

%%time

#################################################################################
# set X
#################################################################################
df_X_to_standardize = df_cleaned_features_IQR

#################################################################################
# Create holdout set and standardize data
#################################################################################
# create hold out set and standardize the data
print('\ncreating hold-out set and standardizing dataset...')
X_train_scaled, X_test_scaled, y_train, y_test = standardize_dataset(df_X_to_standardize, y_cleaned, 0.20, 42, dict_standard_scalers, 'iqr')

#################################################################################
# Visualize the new data
#################################################################################
print('\nDone! Visualizing dataset before and after standardization and outlier transformation...')
try:
    fig, axes = plt.subplots(1, 2, figsize=(12, 5))
    
    # Before scaling
    X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")
    
    # After scaling
    X_train_scaled.plot(kind='kde', ax=axes[1], title="After Scaling")
    
    plt.tight_layout()
    plt.show()
    
    
except:
    pass

#################################################################################
# train the models!
#################################################################################
print('\nDone! Initiating data modeling!')
df_model_evaluation_metrics_iqr, dict_completed_models_iqr, lst_error_model_iqr = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

#################################################################################
# View Model Performance
#################################################################################
print('\nDone! Sorting and saving data...')
# sort models by r2
df_model_evaluation_metrics_iqr = df_model_evaluation_metrics_iqr.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics_iqr = df_model_evaluation_metrics_iqr.reset_index(drop=True, inplace=False)


#save df
df_model_evaluation_metrics_iqr.to_csv('./model_performance_after_outlier_transformation_iqr.csv', index=False)

# view df
#df_model_evaluation_metrics_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.upper, axis=1)
#df_model_evaluation_metrics_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
styled_df = df_model_evaluation_metrics_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
display(HTML(styled_df.to_html()))


#################################################################################
# Sort model data for charting
#################################################################################
print('\nDone! Plotting performance metrics...')
# plot in ascending order
df_model_evaluation_metrics_sorted_iqr = df_model_evaluation_metrics_iqr.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted_iqr = df_model_evaluation_metrics_sorted_iqr.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted_iqr = dict()

lst_sorted_models_iqr = list(df_model_evaluation_metrics_sorted_iqr['model_name'])[::-1]


#################################################################################
# Plot Model Performance
#################################################################################
fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted_iqr['model_name'],  
    x=df_model_evaluation_metrics_sorted_iqr['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()



#################################################################################
# notify
#################################################################################
print('\n\nALL ACTIONS COMPLETE!!!\n\n')

creating hold-out set and standardizing dataset...
X_train: (9397, 8)
y_train: 9397

X_test: (2350, 8)
y_test: 2350

Done! Visualizing dataset before and after standardization and outlier transformation...

Done! Initiating data modeling!
instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:07<04:11,  7.39s/model]

training model ada_boost_regressor complete! 	R2: 0.1822	analysis_duration --> 0:00:07.384942


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:07<00:58,  1.82s/model]

training model ard_regressor complete! 	R2: 0.1853	analysis_duration --> 0:00:00.263121


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

training model bayesian_ridge_regressor complete! 	R2: 0.1853	analysis_duration --> 0:00:00.167477


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

Processing Model:  11%|████▉                                      | 4/35 [00:09<00:55,  1.78s/model]

training model decision_tree_regressor complete! 	R2: 0.3486	analysis_duration --> 0:00:01.705608


######################################################################
# training model --> decision_tree_pruned_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': 5, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

def detect_and_remove_outliers_zscore(df_to_analyze, lst_features, threshold=3):  # Added threshold parameter
    """
    Detects and removes outliers from specified columns in a DataFrame using the Z-score method.

    Args:
        df_to_analyze (pd.DataFrame): The input DataFrame.
        lst_features (list): A list of column names to analyze for outliers.
        threshold (float): The Z-score threshold beyond which a data point is considered an outlier. Default is 3.

    Returns:
        pd.DataFrame: The DataFrame with outliers removed from all specified columns.
        dict: A dictionary where keys are column names and values are lists of outlier indices for each column.
        int: The total number of outlier instances removed.
    """

    all_outlier_indices = {feature: [] for feature in lst_features}
    df_cleaned = df_to_analyze.copy()
    count_total_outliers_removed = 0

    for feature_name in lst_features:
        try:
            df_cleaned[feature_name] = pd.to_numeric(df_cleaned[feature_name], errors='coerce')
            df_cleaned.dropna(subset=[feature_name], inplace=True)

            data = df_cleaned[feature_name].values

            if len(data) < 3:  # Need at least 3 data points for Z-score calculation
                continue

            z_scores = np.abs(zscore(data))  # Calculate absolute Z-scores

            outlier_mask = z_scores > threshold  # Identify outliers based on threshold
            all_outlier_indices[feature_name] = df_cleaned.index[outlier_mask].tolist()

            count_outliers_removed_within_this_feature = len(all_outlier_indices[feature_name])
            print(f'Count outliers to remove from feature: [{feature_name}] --> {count_outliers_removed_within_this_feature}')

        except (TypeError, ValueError) as e:  # Handle potential errors with non-numeric data
            print(f"Error processing column '{feature_name}': {e}")
            continue

    rows_to_remove = []
    for indices in all_outlier_indices.values():
        rows_to_remove.extend(indices)

    rows_to_remove = sorted(list(set(rows_to_remove)))  # Remove duplicates and sort

    count_total_outliers_removed = len(rows_to_remove)

    df_cleaned = df_to_analyze.drop(rows_to_remove, errors='ignore')

    return df_cleaned, all_outlier_indices, count_total_outliers_removed

# NOTE: AI Assisted Code from Gemini

df_cleaned, outlier_indices_dict, count_total_outliers_removed = detect_and_remove_outliers_zscore(df, lst_features, threshold=3)  # You can adjust the threshold
y_cleaned = df_cleaned[DEPENDENT_VAR_NAME]

print(f'\ntotal number of outlier instances removed: --> {count_total_outliers_removed}')

df_cleaned_features_zscore = df_cleaned[lst_features]

Count outliers to remove from feature: [total_volume] --> 171
Count outliers to remove from feature: [4046] --> 181
Count outliers to remove from feature: [4225] --> 180
Count outliers to remove from feature: [4770] --> 259
Count outliers to remove from feature: [total_bags] --> 187
Count outliers to remove from feature: [small_bags] --> 179
Count outliers to remove from feature: [large_bags] --> 251
Count outliers to remove from feature: [xlarge-bags] --> 234

total number of outlier instances removed: --> 467

df_cleaned_features_zscore

###############################################
# print original skewness and kurtosis
###############################################
SKEWNESS = skew(df_X_features)
KURTOSIS = kurtosis(df_X_features)
shape_metric = list(np.sqrt(np.power(SKEWNESS, 2) + np.power(KURTOSIS, 2)))
sum_shape_metric = np.sum(shape_metric)

#print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} \t shape_metric: {shape_metric}')
print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} ')

###############################################
# print transformed skewness and kurtosis
###############################################
SKEWNESS_TRANSFORMED = skew(df_cleaned_features_zscore)
KURTOSIS_TRANSFORMED = kurtosis(df_cleaned_features_zscore)
shape_metric_TRANSFORMED = list(np.sqrt(np.power(SKEWNESS_TRANSFORMED, 2) + np.power(KURTOSIS_TRANSFORMED, 2)))
sum_shape_metric_TRANSFORMED = np.sum(shape_metric_TRANSFORMED)

#print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED} \t shape_metric: {shape_metric_TRANSFORMED}')
print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED}')

################################################
# call fcn - sns kde histplot
################################################
df_to_analyze = df_cleaned_features_zscore
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze = df_cleaned_features_zscore
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

ORIGINAL    sum_shape_metric: 977.3584239619304 
TRANSFORMED sum_shape_metric: 219.25749368112037

%%time

#################################################################################
# set X
#################################################################################
df_X_to_standardize = df_cleaned_features_zscore

#################################################################################
# Create holdout set and standardize data
#################################################################################
# create hold out set and standardize the data
print('\ncreating hold-out set and standardizing dataset...')
X_train_scaled, X_test_scaled, y_train, y_test = standardize_dataset(df_X_to_standardize, y_cleaned, 0.20, 42, dict_standard_scalers, 'zscore')

#################################################################################
# Visualize the new data
#################################################################################
print('\nDone! Visualizing dataset before and after standardization and outlier transformation...')
try:
    fig, axes = plt.subplots(1, 2, figsize=(12, 5))
    
    # Before scaling
    X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")
    
    # After scaling
    X_train_scaled.plot(kind='kde', ax=axes[1], title="After Scaling")
    
    plt.tight_layout()
    plt.show()
    
    
except:
    pass

#################################################################################
# train the models!
#################################################################################
print('\nDone! Initiating data modeling!')
df_model_evaluation_metrics_zscore, dict_completed_models_zscore, lst_error_model_zscore = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

#################################################################################
# View Model Performance
#################################################################################
print('\nDone! Sorting and saving data...')
# sort models by r2
df_model_evaluation_metrics_zscore = df_model_evaluation_metrics_zscore.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics_zscore = df_model_evaluation_metrics_zscore.reset_index(drop=True, inplace=False)


#save df
df_model_evaluation_metrics_zscore.to_csv('./model_performance_after_outlier_transformation_zscore.csv', index=False)

# view df
#df_model_evaluation_metrics_zscore.style.format(precision=3, thousands=",", decimal=".").format_index(str.upper, axis=1)
#df_model_evaluation_metrics_zscore.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
styled_df = df_model_evaluation_metrics_zscore.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
display(HTML(styled_df.to_html()))


#################################################################################
# Sort model data for charting
#################################################################################
print('\nDone! Plotting performance metrics...')
# plot in ascending order
df_model_evaluation_metrics_sorted_zscore = df_model_evaluation_metrics_zscore.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted_zscore = df_model_evaluation_metrics_sorted_zscore.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted_zscore = dict()

lst_sorted_models_zscore = list(df_model_evaluation_metrics_sorted_zscore['model_name'])[::-1]


#################################################################################
# Plot Model Performance
#################################################################################
fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted_zscore['model_name'],  
    x=df_model_evaluation_metrics_sorted_zscore['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()



#################################################################################
# notify
#################################################################################
print('\n\nALL ACTIONS COMPLETE!!!\n\n')

creating hold-out set and standardizing dataset...
X_train: (14225, 8)
y_train: 14225

X_test: (3557, 8)
y_test: 3557

Done! Visualizing dataset before and after standardization and outlier transformation...

Done! Initiating data modeling!
instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:14<08:22, 14.77s/model]

training model ada_boost_regressor complete! 	R2: 0.3267	analysis_duration --> 0:00:14.767303


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   6%|██▍                                        | 2/35 [00:14<03:24,  6.21s/model]

training model ard_regressor complete! 	R2: 0.1236	analysis_duration --> 0:00:00.206992


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:15<01:55,  3.62s/model]

training model bayesian_ridge_regressor complete! 	R2: 0.1242	analysis_duration --> 0:00:00.537284


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

def detect_and_remove_outliers_modified_zscore(df_to_analyze, lst_features, threshold=3.5, verbose=True):
    """
    Detects and removes outliers using the Modified Z-score method.
    """

    all_outlier_indices = {feature: [] for feature in lst_features}
    df_cleaned = df_to_analyze.copy()
    count_total_outliers_removed = 0

    for feature_name in lst_features:
        try:
            df_cleaned[feature_name] = pd.to_numeric(df_cleaned[feature_name], errors='coerce')
            data = df_cleaned[feature_name].dropna().values  # Skip NaNs in calculations

            if len(data) < 3:
                continue

            median_val = np.median(data)
            mad = median_abs_deviation(data)

            if mad == 0:
                mad = 1e-9  # Prevent division by zero

            modified_z_scores = 0.6745 * np.abs(data - median_val) / mad
            outlier_mask = modified_z_scores > threshold

            #CRITICAL FIX: Use the original df index with the mask after dropping NAs.
            original_index_with_no_nans = df_cleaned[feature_name].dropna().index
            outlier_indices = original_index_with_no_nans[outlier_mask].tolist()

            all_outlier_indices[feature_name] = outlier_indices

            if verbose:
                print(f'Feature [{feature_name}] - Outliers detected: {len(outlier_indices)}')

        except (TypeError, ValueError) as e:
            print(f"Error processing column '{feature_name}': {e}")
            continue

    # Collect unique outlier row indices
    rows_to_remove = list(set(idx for indices in all_outlier_indices.values() for idx in indices))
    count_total_outliers_removed = len(rows_to_remove)

    # Drop rows from the original DataFrame
    df_cleaned = df_to_analyze.drop(index=rows_to_remove, errors='ignore')

    return df_cleaned, all_outlier_indices, count_total_outliers_removed

#NOTE: AI Assisted Code from Gemini

# Example usage:
df_cleaned, outlier_indices_dict, count_total_outliers_removed = detect_and_remove_outliers_modified_zscore(df, lst_features, threshold=3.5, verbose=True)
y_cleaned = df_cleaned[DEPENDENT_VAR_NAME]

print(f'\nTotal outliers removed: {count_total_outliers_removed}')
df_cleaned_features_modified_zscore_mad = df_cleaned[lst_features]

Feature [total_volume] - Outliers detected: 3503
Feature [4046] - Outliers detected: 6322
Feature [4225] - Outliers detected: 4225
Feature [4770] - Outliers detected: 6825
Feature [total_bags] - Outliers detected: 2715
Feature [small_bags] - Outliers detected: 2755
Feature [large_bags] - Outliers detected: 5192
Feature [xlarge-bags] - Outliers detected: 6201

Total outliers removed: 9744

df_cleaned_features_modified_zscore_mad

###############################################
# print original skewness and kurtosis
###############################################
SKEWNESS = skew(df_X_features)
KURTOSIS = kurtosis(df_X_features)
shape_metric = list(np.sqrt(np.power(SKEWNESS, 2) + np.power(KURTOSIS, 2)))
sum_shape_metric = np.sum(shape_metric)

#print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} \t shape_metric: {shape_metric}')
print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} ')

###############################################
# print transformed skewness and kurtosis
###############################################
SKEWNESS_TRANSFORMED = skew(df_cleaned_features_modified_zscore_mad)
KURTOSIS_TRANSFORMED = kurtosis(df_cleaned_features_modified_zscore_mad)
shape_metric_TRANSFORMED = list(np.sqrt(np.power(SKEWNESS_TRANSFORMED, 2) + np.power(KURTOSIS_TRANSFORMED, 2)))
sum_shape_metric_TRANSFORMED = np.sum(shape_metric_TRANSFORMED)

#print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED} \t shape_metric: {shape_metric_TRANSFORMED}')
print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED}')

################################################
# call fcn - sns kde histplot
################################################
df_to_analyze = df_cleaned_features_modified_zscore_mad
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze = df_cleaned_features_modified_zscore_mad
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

ORIGINAL    sum_shape_metric: 977.3584239619304 
TRANSFORMED sum_shape_metric: nan

%%time

#################################################################################
# set X
#################################################################################
df_X_to_standardize = df_cleaned_features_modified_zscore_mad

#################################################################################
# Create holdout set and standardize data
#################################################################################
# create hold out set and standardize the data
print('\ncreating hold-out set and standardizing dataset...')
X_train_scaled, X_test_scaled, y_train, y_test = standardize_dataset(df_X_to_standardize, y_cleaned, 0.20, 42, dict_standard_scalers, 'modified_zscore_mad')

#################################################################################
# Visualize the new data
#################################################################################
print('\nDone! Visualizing dataset before and after standardization and outlier transformation...')
try:
    fig, axes = plt.subplots(1, 2, figsize=(12, 5))
    
    # Before scaling
    X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")
    
    # After scaling
    X_train_scaled.plot(kind='kde', ax=axes[1], title="After Scaling")
    
    plt.tight_layout()
    plt.show()
    
    
except:
    pass

#################################################################################
# train the models!
#################################################################################
print('\nDone! Initiating data modeling!')
df_model_evaluation_metrics_modified_zscore_mad, dict_completed_models_modified_zscore_mad, lst_error_model_modified_zscore_mad = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

#################################################################################
# View Model Performance
#################################################################################
print('\nDone! Sorting and saving data...')
# sort models by r2
df_model_evaluation_metrics_modified_zscore_mad = df_model_evaluation_metrics_modified_zscore_mad.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics_modified_zscore_mad = df_model_evaluation_metrics_modified_zscore_mad.reset_index(drop=True, inplace=False)


#save df
df_model_evaluation_metrics_modified_zscore_mad.to_csv('./model_performance_after_outlier_transformation_modified_zscore_mad.csv', index=False)

# view df
#df_model_evaluation_metrics_modified_zscore_mad.style.format(precision=3, thousands=",", decimal=".").format_index(str.upper, axis=1)
#df_model_evaluation_metrics_modified_zscore_mad.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
styled_df = df_model_evaluation_metrics_modified_zscore_mad.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
display(HTML(styled_df.to_html()))


#################################################################################
# Sort model data for charting
#################################################################################
print('\nDone! Plotting performance metrics...')
# plot in ascending order
df_model_evaluation_metrics_sorted_modified_zscore_mad = df_model_evaluation_metrics_modified_zscore_mad.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted_modified_zscore_mad = df_model_evaluation_metrics_sorted_modified_zscore_mad.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted_modified_zscore_mad = dict()

lst_sorted_models_modified_zscore_mad = list(df_model_evaluation_metrics_sorted_modified_zscore_mad['model_name'])[::-1]


#################################################################################
# Plot Model Performance
#################################################################################
fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted_modified_zscore_mad['model_name'],  
    x=df_model_evaluation_metrics_sorted_modified_zscore_mad['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()



#################################################################################
# notify
#################################################################################
print('\n\nALL ACTIONS COMPLETE!!!\n\n')

creating hold-out set and standardizing dataset...
X_train: (6804, 8)
y_train: 6804

X_test: (1701, 8)
y_test: 1701

Done! Visualizing dataset before and after standardization and outlier transformation...

Done! Initiating data modeling!
instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:04<02:27,  4.33s/model]

training model ada_boost_regressor complete! 	R2: 0.1550	analysis_duration --> 0:00:04.324284


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   6%|██▍                                        | 2/35 [00:04<01:04,  1.97s/model]

training model ard_regressor complete! 	R2: 0.0933	analysis_duration --> 0:00:00.306224


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:04<00:37,  1.17s/model]

training model bayesian_ridge_regressor complete! 	R2: 0.0927	analysis_duration --> 0:00:00.195386


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

Processing Model:  11%|████▉                                      | 4/35 [00:05<00:33,  1.08s/model]

def detect_and_remove_outliers_dbscan(df_to_analyze, lst_features, epsilon=0.5, min_samples=5, verbose=True, override_feature_to_remove_outliers='xlarge-bags'):
    """Detects and removes outliers using DBSCAN (without scaling)."""

    all_outlier_indices = {feature: [] for feature in lst_features}
    df_cleaned = df_to_analyze.copy()
    count_total_outliers_removed = 0

    for feature_name in lst_features:
        try:
            df_cleaned[feature_name] = pd.to_numeric(df_cleaned[feature_name], errors='coerce')

            # 1. Handle NaNs: Create a mask of NON-NaN values
            nan_mask = df_cleaned[feature_name].notna()

            # 2. Extract data *and* indices for DBSCAN based on the nan_mask
            data = df_cleaned.loc[nan_mask, feature_name].values.reshape(-1, 1)
            original_indices = df_cleaned.loc[nan_mask].index

            if len(data) < min_samples + 1:
                if verbose:
                    print(f"Feature [{feature_name}] skipped due to insufficient data points (< {min_samples + 1}).")
                continue

            # 3. Perform DBSCAN
            dbscan = DBSCAN(eps=epsilon, min_samples=min_samples)
            clusters = dbscan.fit_predict(data)

            # 4. Identify outliers
            outlier_mask = clusters == -1

            # 5. Get ORIGINAL indices of outliers
            outlier_indices_current_feature = original_indices[outlier_mask].tolist()

            if override_feature_to_remove_outliers is not None and feature_name == override_feature_to_remove_outliers:
                all_outlier_indices[feature_name] = outlier_indices_current_feature

            if verbose:
                print(f'Feature [{feature_name}] - Outliers detected: {len(outlier_indices_current_feature)}')

        except (TypeError, ValueError) as e:
            print(f"Error processing column '{feature_name}': {e}")
            continue

    # 6. Collect and remove duplicate outlier indices
    rows_to_remove = list(set(idx for indices in all_outlier_indices.values() for idx in indices))
    count_total_outliers_removed = len(rows_to_remove)

    if count_total_outliers_removed == 0 and verbose:
        print("\nNo outliers detected. No rows removed.")

    df_cleaned = df_to_analyze.drop(index=rows_to_remove, errors='ignore')  # Remove from ORIGINAL DataFrame

    return df_cleaned, all_outlier_indices, count_total_outliers_removed

# Example usage:
df_cleaned, outlier_indices_dict, count_total_outliers_removed = detect_and_remove_outliers_dbscan(df, lst_features, epsilon=0.5, min_samples=5, verbose=True, override_feature_to_remove_outliers='xlarge-bags')
y_cleaned = df_cleaned[DEPENDENT_VAR_NAME]

print(f'\nTotal outliers removed: {count_total_outliers_removed}')
df_cleaned_features_dbscan = df_cleaned[lst_features]

Feature [total_volume] - Outliers detected: 18223
Feature [4046] - Outliers detected: 14257
Feature [4225] - Outliers detected: 16562
Feature [4770] - Outliers detected: 7919
Feature [total_bags] - Outliers detected: 17939
Feature [small_bags] - Outliers detected: 16715
Feature [large_bags] - Outliers detected: 12141
Feature [xlarge-bags] - Outliers detected: 4206

Total outliers removed: 4206

df_cleaned_features_dbscan

###############################################
# print original skewness and kurtosis
###############################################
SKEWNESS = skew(df_X_features)
KURTOSIS = kurtosis(df_X_features)
shape_metric = list(np.sqrt(np.power(SKEWNESS, 2) + np.power(KURTOSIS, 2)))
sum_shape_metric = np.sum(shape_metric)

#print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} \t shape_metric: {shape_metric}')
print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} ')

###############################################
# print transformed skewness and kurtosis
###############################################
SKEWNESS_TRANSFORMED = skew(df_cleaned_features_dbscan)
KURTOSIS_TRANSFORMED = kurtosis(df_cleaned_features_dbscan)
shape_metric_TRANSFORMED = list(np.sqrt(np.power(SKEWNESS_TRANSFORMED, 2) + np.power(KURTOSIS_TRANSFORMED, 2)))
sum_shape_metric_TRANSFORMED = np.sum(shape_metric_TRANSFORMED)

#print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED} \t shape_metric: {shape_metric_TRANSFORMED}')
print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED}')

################################################
# call fcn - sns kde histplot
################################################
df_to_analyze = df_cleaned_features_dbscan
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze = df_cleaned_features_dbscan
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

ORIGINAL    sum_shape_metric: 977.3584239619304 
TRANSFORMED sum_shape_metric: 645.0474951320733

%%time

#################################################################################
# set X
#################################################################################
df_X_to_standardize = df_cleaned_features_dbscan

#################################################################################
# Create holdout set and standardize data
#################################################################################
# create hold out set and standardize the data
print('\ncreating hold-out set and standardizing dataset...')
X_train_scaled, X_test_scaled, y_train, y_test = standardize_dataset(df_X_to_standardize, y_cleaned, 0.20, 42, dict_standard_scalers, 'dbscan')

#################################################################################
# Visualize the new data
#################################################################################
print('\nDone! Visualizing dataset before and after standardization and outlier transformation...')
try:
    fig, axes = plt.subplots(1, 2, figsize=(12, 5))
    
    # Before scaling
    X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")
    
    # After scaling
    X_train_scaled.plot(kind='kde', ax=axes[1], title="After Scaling")
    
    plt.tight_layout()
    plt.show()
    
    
except:
    pass

#################################################################################
# train the models!
#################################################################################
print('\nDone! Initiating data modeling!')
df_model_evaluation_metrics_dbscan, dict_completed_models_dbscan, lst_error_model_dbscan = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

#################################################################################
# View Model Performance
#################################################################################
print('\nDone! Sorting and saving data...')
# sort models by r2
df_model_evaluation_metrics_dbscan = df_model_evaluation_metrics_dbscan.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics_dbscan = df_model_evaluation_metrics_dbscan.reset_index(drop=True, inplace=False)


#save df
df_model_evaluation_metrics_dbscan.to_csv('./model_performance_after_outlier_transformation_dbscan.csv', index=False)

# view df
#df_model_evaluation_metrics_dbscan.style.format(precision=3, thousands=",", decimal=".").format_index(str.upper, axis=1)
#df_model_evaluation_metrics_dbscan.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
styled_df = df_model_evaluation_metrics_dbscan.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
display(HTML(styled_df.to_html()))


#################################################################################
# Sort model data for charting
#################################################################################
print('\nDone! Plotting performance metrics...')
# plot in ascending order
df_model_evaluation_metrics_sorted_dbscan = df_model_evaluation_metrics_dbscan.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted_dbscan = df_model_evaluation_metrics_sorted_dbscan.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted_dbscan = dict()

lst_sorted_models_dbscan = list(df_model_evaluation_metrics_sorted_dbscan['model_name'])[::-1]


#################################################################################
# Plot Model Performance
#################################################################################
fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted_dbscan['model_name'],  
    x=df_model_evaluation_metrics_sorted_dbscan['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()



#################################################################################
# notify
#################################################################################
print('\n\nALL ACTIONS COMPLETE!!!\n\n')

creating hold-out set and standardizing dataset...
X_train: (11234, 8)
y_train: 11234

X_test: (2809, 8)
y_test: 2809

Done! Visualizing dataset before and after standardization and outlier transformation...

Done! Initiating data modeling!
instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:06<03:29,  6.15s/model]

training model ada_boost_regressor complete! 	R2: 0.4012	analysis_duration --> 0:00:06.150319


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   6%|██▍                                        | 2/35 [00:06<01:30,  2.73s/model]

training model ard_regressor complete! 	R2: 0.1465	analysis_duration --> 0:00:00.326876


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:06<00:50,  1.59s/model]

training model bayesian_ridge_regressor complete! 	R2: 0.1466	analysis_duration --> 0:00:00.211515


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

def clip_outliers_iqr(df_to_analyze, lst_features):
    """
    Clips outliers to the IQR bounds instead of removing them.

    Args:
        df_to_analyze (pd.DataFrame): The input DataFrame.
        lst_features (list): A list of column names to analyze for outliers.

    Returns:
        pd.DataFrame: The DataFrame with outliers clipped to the IQR bounds.
        dict: A dictionary where keys are column names and values are lists of outlier indices for each column.
    """

    all_outlier_indices = {feature: [] for feature in lst_features}
    df_clipped = df_to_analyze.copy()  # Work on a copy

    for feature_name in lst_features:
        try:
            df_clipped[feature_name] = pd.to_numeric(df_clipped[feature_name], errors='coerce')  # Convert to numeric
            df_clipped.dropna(subset=[feature_name], inplace=True) # Drop rows with NaN in the column

            data = df_clipped[feature_name].values

            if len(data) < 3:
                continue

            q1 = np.percentile(data, 25)
            q3 = np.percentile(data, 75)
            iqr = q3 - q1

            lower_bound = q1 - 1.5 * iqr
            upper_bound = q3 + 1.5 * iqr

            # Clip values:
            df_clipped[feature_name] = np.clip(df_clipped[feature_name], lower_bound, upper_bound)

            # Identify outliers (for informational purposes) - this does not affect the clipping
            outlier_mask = (data < lower_bound) | (data > upper_bound)
            all_outlier_indices[feature_name] = df_clipped.index[outlier_mask].tolist()

            count_outliers_identified = len(all_outlier_indices[feature_name])
            print(f'Outliers identified in feature [{feature_name}]: {count_outliers_identified}')


        except (TypeError, ValueError) as e:
            print(f"Error processing column '{feature_name}': {e}")
            continue

    return df_clipped, all_outlier_indices

# Example usage:
df_cleaned, outlier_indices_dict = clip_outliers_iqr(df, lst_features)
y_cleaned = df_cleaned[DEPENDENT_VAR_NAME]


print(f'\nTotal outliers clipped: {count_total_outliers_removed}')
df_cleaned_features_clipped_iqr = df_cleaned[lst_features]

Outliers identified in feature [total_volume]: 2297
Outliers identified in feature [4046]: 2518
Outliers identified in feature [4225]: 2519
Outliers identified in feature [4770]: 2918
Outliers identified in feature [total_bags]: 2533
Outliers identified in feature [small_bags]: 2407
Outliers identified in feature [large_bags]: 2707
Outliers identified in feature [xlarge-bags]: 4014

Total outliers clipped: 4206

df_cleaned_features_clipped_iqr

###############################################
# print original skewness and kurtosis
###############################################
SKEWNESS = skew(df_X_features)
KURTOSIS = kurtosis(df_X_features)
shape_metric = list(np.sqrt(np.power(SKEWNESS, 2) + np.power(KURTOSIS, 2)))
sum_shape_metric = np.sum(shape_metric)

#print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} \t shape_metric: {shape_metric}')
print(f'ORIGINAL    sum_shape_metric: {sum_shape_metric} ')

###############################################
# print transformed skewness and kurtosis
###############################################
SKEWNESS_TRANSFORMED = skew(df_cleaned_features_clipped_iqr)
KURTOSIS_TRANSFORMED = kurtosis(df_cleaned_features_clipped_iqr)
shape_metric_TRANSFORMED = list(np.sqrt(np.power(SKEWNESS_TRANSFORMED, 2) + np.power(KURTOSIS_TRANSFORMED, 2)))
sum_shape_metric_TRANSFORMED = np.sum(shape_metric_TRANSFORMED)

#print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED} \t shape_metric: {shape_metric_TRANSFORMED}')
print(f'TRANSFORMED sum_shape_metric: {sum_shape_metric_TRANSFORMED}')

################################################
# call fcn - sns kde histplot
################################################
df_to_analyze = df_cleaned_features_clipped_iqr
plot_kde_histplot(df_to_analyze, False, 3, 'Avocado Distibution of each Feature')

################################################
# call fcn - sns boxplot
################################################
#df_to_analyze = df.select_dtypes(include=[np.number])
df_to_analyze = df_cleaned_features_clipped_iqr
plot_boxplot(df_to_analyze, "Boxplot of Attributes", False)

ORIGINAL    sum_shape_metric: 977.3584239619304 
TRANSFORMED sum_shape_metric: 9.802148593703722

%%time

#################################################################################
# set X
#################################################################################
df_X_to_standardize = df_cleaned_features_clipped_iqr

#################################################################################
# Create holdout set and standardize data
#################################################################################
# create hold out set and standardize the data
print('\ncreating hold-out set and standardizing dataset...')
X_train_scaled, X_test_scaled, y_train, y_test = standardize_dataset(df_X_to_standardize, y_cleaned, 0.20, 42, dict_standard_scalers, 'clipped_iqr')

#################################################################################
# Visualize the new data
#################################################################################
print('\nDone! Visualizing dataset before and after standardization and outlier transformation...')
try:
    fig, axes = plt.subplots(1, 2, figsize=(12, 5))
    
    # Before scaling
    X_train.plot(kind='kde', ax=axes[0], title="Before Scaling")
    
    # After scaling
    X_train_scaled.plot(kind='kde', ax=axes[1], title="After Scaling")
    
    plt.tight_layout()
    plt.show()
    
    
except:
    pass

#################################################################################
# train the models!
#################################################################################
print('\nDone! Initiating data modeling!')
df_model_evaluation_metrics_clipped_iqr, dict_completed_models_clipped_iqr, lst_error_model_clipped_iqr = instantiate_and_model_data(X_train_scaled, X_test_scaled, y_train, y_test)

#################################################################################
# View Model Performance
#################################################################################
print('\nDone! Sorting and saving data...')
# sort models by r2
df_model_evaluation_metrics_clipped_iqr = df_model_evaluation_metrics_clipped_iqr.sort_values(by='r2', ascending=False, inplace=False)
df_model_evaluation_metrics_clipped_iqr = df_model_evaluation_metrics_clipped_iqr.reset_index(drop=True, inplace=False)


#save df
df_model_evaluation_metrics_clipped_iqr.to_csv('./model_performance_after_outlier_transformation_clipped_iqr.csv', index=False)

# view df
#df_model_evaluation_metrics_clipped_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.upper, axis=1)
#df_model_evaluation_metrics_clipped_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
styled_df = df_model_evaluation_metrics_clipped_iqr.style.format(precision=3, thousands=",", decimal=".").format_index(str.lower, axis=1)
display(HTML(styled_df.to_html()))


#################################################################################
# Sort model data for charting
#################################################################################
print('\nDone! Plotting performance metrics...')
# plot in ascending order
df_model_evaluation_metrics_sorted_clipped_iqr = df_model_evaluation_metrics_clipped_iqr.sort_values(by='r2', ascending=True, inplace=False)
df_model_evaluation_metrics_sorted_clipped_iqr = df_model_evaluation_metrics_sorted_clipped_iqr.reset_index(drop=True, inplace=False)

# soft dict of models for displaying feature importance below
dict_completed_models_sorted_clipped_iqr = dict()

lst_sorted_models_clipped_iqr = list(df_model_evaluation_metrics_sorted_clipped_iqr['model_name'])[::-1]


#################################################################################
# Plot Model Performance
#################################################################################
fig = go.Figure()

fig.add_trace(go.Bar(
    y=df_model_evaluation_metrics_sorted_clipped_iqr['model_name'],  
    x=df_model_evaluation_metrics_sorted_clipped_iqr['r2'],  
    orientation='h',  # Horizontal bar chart
    marker=dict(color='blue', line=dict(color='black', width=1)),  
))

# Update layout with dashed vertical gridlines
fig.update_layout(
    title="Model Performance Based on R-squared",
    xaxis_title="R-squared",
    yaxis_title="Model Name",
    xaxis=dict(
        showgrid=True, 
        zeroline=True,
        gridcolor='lightgray',  
        gridwidth=0.5,      
        griddash='dash'      
    ),
    yaxis=dict(tickmode='linear'),
    height=900, width=1300,  
    plot_bgcolor="white",  
    paper_bgcolor="white"   
)

# Show plot
fig.show()



#################################################################################
# notify
#################################################################################
print('\n\nALL ACTIONS COMPLETE!!!\n\n')

creating hold-out set and standardizing dataset...
X_train: (14599, 8)
y_train: 14599

X_test: (3650, 8)
y_test: 3650

Done! Visualizing dataset before and after standardization and outlier transformation...

Done! Initiating data modeling!
instantiating models...
model instantiation complete!

Processing Model:   0%|                                                   | 0/35 [00:00<?, ?model/s]


######################################################################
# training model --> ada_boost_regressor
######################################################################
hyperparameters --> {'estimator': None, 'learning_rate': 1.0, 'loss': 'linear', 'n_estimators': 50, 'random_state': 42}

Processing Model:   3%|█▏                                         | 1/35 [00:08<04:38,  8.19s/model]

training model ada_boost_regressor complete! 	R2: 0.3454	analysis_duration --> 0:00:08.184067


######################################################################
# training model --> ard_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'max_iter': 300, 'threshold_lambda': 10000.0, 'tol': 0.001, 'verbose': False}

Processing Model:   6%|██▍                                        | 2/35 [00:08<02:00,  3.65s/model]

training model ard_regressor complete! 	R2: 0.2969	analysis_duration --> 0:00:00.468076


######################################################################
# training model --> bayesian_ridge_regressor
######################################################################
hyperparameters --> {'alpha_1': 1e-06, 'alpha_2': 1e-06, 'alpha_init': None, 'compute_score': False, 'copy_X': True, 'fit_intercept': True, 'lambda_1': 1e-06, 'lambda_2': 1e-06, 'lambda_init': None, 'max_iter': 300, 'tol': 0.001, 'verbose': False}

Processing Model:   9%|███▋                                       | 3/35 [00:08<01:07,  2.12s/model]

training model bayesian_ridge_regressor complete! 	R2: 0.2967	analysis_duration --> 0:00:00.275674


######################################################################
# training model --> decision_tree_regressor
######################################################################
hyperparameters --> {'ccp_alpha': 0.0, 'criterion': 'squared_error', 'max_depth': None, 'max_features': None, 'max_leaf_nodes': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'monotonic_cst': None, 'random_state': 42, 'splitter': 'best'}

def get_model(df_model_evaluation_metrics, dict_completed_models, outlier_transformation_description, dict_top_models):
    try:
        top_model_name = df_model_evaluation_metrics.iloc[0].model_name
        top_model_r2 = df_model_evaluation_metrics.iloc[0].r2
        top_model = dict_completed_models[top_model_name]

        # get dataframe of only top model's metrics
        df_model_evaluation_metrics = df_model_evaluation_metrics[df_model_evaluation_metrics['model_name'] == top_model_name]

        # add outlier transformation description
        df_model_evaluation_metrics.insert(0, 'outlier_transformation', outlier_transformation_description)

        # store the top model for future reference
        dict_top_models[outlier_transformation_description] = top_model

        return top_model_name, top_model_r2, top_model, df_model_evaluation_metrics
    except Exception as e:
        print('\n\n* * * EXCEPTION caught in get_model mtd')
        print(f"Exception message: {e}")  # Print the exception message
        traceback.print_exc()  # Print the full stack trace

# create dictionary of top models
dict_top_models = dict()

# get top model before outlier transformations
top_model_name_before_outlier_transformation, top_model_r2_before_outlier_transformation, top_model_before_outlier_transformation, df_top_model_evaluation_metrics_before_outlier_transformation = get_model(df_model_evaluation_metrics, dict_completed_models, 'before_transformation', dict_top_models)

# get top model _iqr transformation
top_model_name_iqr_outlier_transformation, top_model_r2_iqr_outlier_transformation, top_model_iqr_outlier_transformation, df_top_model_evaluation_metrics_iqr_outlier_transformation = get_model(df_model_evaluation_metrics_iqr, dict_completed_models_iqr, 'iqr', dict_top_models)

# get top model _zscore transformation
top_model_name_zscore_outlier_transformation, top_model_r2_zscore_outlier_transformation, top_model_zscore_outlier_transformation, df_top_model_evaluation_metrics_zscore_outlier_transformation = get_model(df_model_evaluation_metrics_zscore, dict_completed_models_zscore, 'zscore', dict_top_models)

# get top model _modified_zscore_mad transformation
top_model_name_modified_zscore_mad_outlier_transformation, top_model_r2_modified_zscore_mad_outlier_transformation, top_model_modified_zscore_mad_outlier_transformation, df_top_model_evaluation_metrics_zscore_mad_outlier_transformation = get_model(df_model_evaluation_metrics_modified_zscore_mad, dict_completed_models_modified_zscore_mad, 'modified_zscore_mad', dict_top_models)

# get top model _dbscan transformation
top_model_name_dbscan_outlier_transformation, top_model_r2_dbscan_outlier_transformation, top_model_dbscan_outlier_transformation, df_top_model_evaluation_metrics_dbscan_outlier_transformation = get_model(df_model_evaluation_metrics_dbscan, dict_completed_models_dbscan, 'dbscan', dict_top_models)

# get top model _clipped_iqr transformation
top_model_name_clipped_iqr_outlier_transformation, top_model_r2_clipped_iqr_outlier_transformation, top_model_clipped_iqr_outlier_transformation, df_top_model_evaluation_metrics_clipped_iqr_outlier_transformation = get_model(df_model_evaluation_metrics_clipped_iqr, dict_completed_models_clipped_iqr, 'clipped_iqr', dict_top_models)



# display stats
print('top_model_name\t\t r2\t\t model\t\t\t\t\t model_description')
print(top_model_name_before_outlier_transformation, '\t', top_model_r2_before_outlier_transformation, '\t', top_model_before_outlier_transformation, '\t', 'before_outlier_transformation')
print(top_model_name_iqr_outlier_transformation, '\t', top_model_r2_iqr_outlier_transformation, '\t', top_model_iqr_outlier_transformation, '\t', 'iqr_outlier_transformation')
print(top_model_name_zscore_outlier_transformation, '\t', top_model_r2_zscore_outlier_transformation, '\t', top_model_zscore_outlier_transformation, '\t', 'zscore_outlier_transformation')
print(top_model_name_modified_zscore_mad_outlier_transformation, '\t', top_model_r2_modified_zscore_mad_outlier_transformation, '\t', top_model_modified_zscore_mad_outlier_transformation, '\t', 'modified_zscore_mad_outlier_transformation')
print(top_model_name_dbscan_outlier_transformation, '\t', top_model_r2_dbscan_outlier_transformation, '\t', top_model_dbscan_outlier_transformation, '\t', 'dbscan_outlier_transformation')
print(top_model_name_clipped_iqr_outlier_transformation, '\t', top_model_r2_clipped_iqr_outlier_transformation, '\t', top_model_clipped_iqr_outlier_transformation, '\t', 'clipped_iqr_outlier_transformation')

print('\ndict_top_mdels entries:')
print('='*40)
for key in dict_top_models.keys():
    print(f'{key}, {dict_top_models[key]}')

top_model_name		 r2		 model					 model_description
extra_trees_regressor 	 0.7657 	 ExtraTreesRegressor(random_state=42) 	 before_outlier_transformation
extra_trees_regressor 	 0.6825 	 ExtraTreesRegressor(random_state=42) 	 iqr_outlier_transformation
extra_trees_regressor 	 0.7732 	 ExtraTreesRegressor(random_state=42) 	 zscore_outlier_transformation
extra_trees_regressor 	 0.5823 	 ExtraTreesRegressor(random_state=42) 	 modified_zscore_mad_outlier_transformation
extra_trees_regressor 	 0.7288 	 ExtraTreesRegressor(random_state=42) 	 dbscan_outlier_transformation
extra_trees_regressor 	 0.7349 	 ExtraTreesRegressor(random_state=42) 	 clipped_iqr_outlier_transformation

dict_top_mdels entries:
========================================
before_transformation, ExtraTreesRegressor(random_state=42)
iqr, ExtraTreesRegressor(random_state=42)
zscore, ExtraTreesRegressor(random_state=42)
modified_zscore_mad, ExtraTreesRegressor(random_state=42)
dbscan, ExtraTreesRegressor(random_state=42)
clipped_iqr, ExtraTreesRegressor(random_state=42)

# first create the new dataframe
df_top_model_performance_metrics_combined = pd.DataFrame(columns=df_top_model_evaluation_metrics_before_outlier_transformation.columns)

# merge dataframes transformation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_before_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_iqr_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_zscore_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_zscore_mad_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_dbscan_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation
df_top_model_performance_metrics_combined = pd.concat([df_top_model_performance_metrics_combined, df_top_model_evaluation_metrics_clipped_iqr_outlier_transformation], axis=0)  # axis=1 for column-wise concatenation, # axis=0 for row-wise concatenation

# sort models by r2
df_top_model_performance_metrics_combined = df_top_model_performance_metrics_combined.sort_values(by='r2', ascending=False, inplace=False)
df_top_model_performance_metrics_combined = df_top_model_performance_metrics_combined.reset_index(drop=True, inplace=False)

#save df
df_top_model_performance_metrics_combined.to_csv('./top_model_performance_combined_from_outlier_transformation.csv', index=False)

# print final results
print_df_styled_widget(df_top_model_performance_metrics_combined, 4)

# create single dataframe of best model
df_best_model = df_top_model_performance_metrics_combined.iloc[[0]]
print_df_styled_widget(df_best_model, 4)

# select best model
best_model_name = df_top_model_performance_metrics_combined.iloc[0].model_name
best_model_outlier_transformation = df_top_model_performance_metrics_combined.iloc[0].outlier_transformation
best_model_r2 = df_top_model_performance_metrics_combined.iloc[0].r2
best_model = dict_top_models[best_model_outlier_transformation]
best_model_scaler = dict_standard_scalers[best_model_outlier_transformation]

# notify
print(f'Best model --> {best_model_name} \noutlier_transformation --> {best_model_outlier_transformation} \nR2 --> {best_model_r2} \nmodel --> {best_model} \nmodel scaler --> {best_model_scaler}\n\n')

# view feature importance
lst_coefficients, plt = display_feature_importance_chart(best_model, best_model_name, list(X_train.columns), 10, '_feature_importance_best_model_' + str(best_model_outlier_transformation) + '_' + str(best_model_name), X_test_scaled, y_test, True)

Best model --> extra_trees_regressor 
outlier_transformation --> zscore 
R2 --> 0.7732 
model --> ExtraTreesRegressor(random_state=42) 
model scaler --> StandardScaler()


coefficients:
====================
4046	0.19192857052505133
total_volume	0.1747568197359546
4225	0.1418204317398782
large_bags	0.1285193360399352
total_bags	0.12747147124753355
small_bags	0.10640741366666613
4770	0.09290926895938907
xlarge-bags	0.03618668808559197

# get instance from X_test or format instance in the same format as data arranged in y_test
# note, use the same scaler to standardize the instance (disregard if taking from X_test and if X_test was already standardized)

# get instance (as a series) and then flatten the values from the instance just to get the data as an array ready for modeling
# here, we'll just take an index from the original dataframe of features - feel free to change this index especially to ensure
# it is an index not included in the training set
index = 0
instance = np.array(df_X_features.iloc[index]).reshape(1, -1)

# standardize the instance
instance_scaled = best_model_scaler.transform(instance) 

# get model's prediction
prediction = best_model.predict(instance_scaled) 

# notify
print(f'Avacado Price Prediction --> {round(prediction[0], 2)}')

print(f'Actual Average Avacado Price --> {df.iloc[index].average_price}')

Avacado Price Prediction --> 1.31
Actual Average Avacado Price --> 1.33

	Unnamed: 0	Date	AveragePrice	Total Volume	4046	4225	4770	Total Bags	Small Bags	Large Bags	XLarge Bags	type	year	region
0	0	2015-12-27	1.33	64236.62	1036.74	54454.85	48.16	8696.87	8603.62	93.25	0.0	conventional	2015	Albany
1	1	2015-12-20	1.35	54876.98	674.28	44638.81	58.33	9505.56	9408.07	97.49	0.0	conventional	2015	Albany
2	2	2015-12-13	0.93	118220.22	794.70	109149.67	130.50	8145.35	8042.21	103.14	0.0	conventional	2015	Albany
3	3	2015-12-06	1.08	78992.15	1132.00	71976.41	72.58	5811.16	5677.40	133.76	0.0	conventional	2015	Albany
4	4	2015-11-29	1.28	51039.60	941.48	43838.39	75.78	6183.95	5986.26	197.69	0.0	conventional	2015	Albany
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
18244	7	2018-02-04	1.63	17074.83	2046.96	1529.20	0.00	13498.67	13066.82	431.85	0.0	organic	2018	WestTexNewMexico
18245	8	2018-01-28	1.71	13888.04	1191.70	3431.50	0.00	9264.84	8940.04	324.80	0.0	organic	2018	WestTexNewMexico
18246	9	2018-01-21	1.87	13766.76	1191.92	2452.79	727.94	9394.11	9351.80	42.31	0.0	organic	2018	WestTexNewMexico
18247	10	2018-01-14	1.93	16205.22	1527.63	2981.04	727.01	10969.54	10919.54	50.00	0.0	organic	2018	WestTexNewMexico
18248	11	2018-01-07	1.62	17489.58	2894.77	2356.13	224.53	12014.15	11988.14	26.01	0.0	organic	2018	WestTexNewMexico

	unnamed_0	date	average_price	total_volume	4046	4225	4770	total_bags	small_bags	large_bags	xlarge-bags	type	year	region
0	0	2015-12-27	1.33	64236.62	1036.74	54454.85	48.16	8696.87	8603.62	93.25	0.0	conventional	2015	Albany
1	1	2015-12-20	1.35	54876.98	674.28	44638.81	58.33	9505.56	9408.07	97.49	0.0	conventional	2015	Albany
2	2	2015-12-13	0.93	118220.22	794.70	109149.67	130.50	8145.35	8042.21	103.14	0.0	conventional	2015	Albany
3	3	2015-12-06	1.08	78992.15	1132.00	71976.41	72.58	5811.16	5677.40	133.76	0.0	conventional	2015	Albany
4	4	2015-11-29	1.28	51039.60	941.48	43838.39	75.78	6183.95	5986.26	197.69	0.0	conventional	2015	Albany
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
18244	7	2018-02-04	1.63	17074.83	2046.96	1529.20	0.00	13498.67	13066.82	431.85	0.0	organic	2018	WestTexNewMexico
18245	8	2018-01-28	1.71	13888.04	1191.70	3431.50	0.00	9264.84	8940.04	324.80	0.0	organic	2018	WestTexNewMexico
18246	9	2018-01-21	1.87	13766.76	1191.92	2452.79	727.94	9394.11	9351.80	42.31	0.0	organic	2018	WestTexNewMexico
18247	10	2018-01-14	1.93	16205.22	1527.63	2981.04	727.01	10969.54	10919.54	50.00	0.0	organic	2018	WestTexNewMexico
18248	11	2018-01-07	1.62	17489.58	2894.77	2356.13	224.53	12014.15	11988.14	26.01	0.0	organic	2018	WestTexNewMexico

	date	average_price	total_volume	4046	4225	4770	total_bags	small_bags	large_bags	xlarge-bags	type	year	region
0	2015-12-27	1.33	64236.62	1036.74	54454.85	48.16	8696.87	8603.62	93.25	0.0	conventional	2015	Albany
1	2015-12-20	1.35	54876.98	674.28	44638.81	58.33	9505.56	9408.07	97.49	0.0	conventional	2015	Albany
2	2015-12-13	0.93	118220.22	794.70	109149.67	130.50	8145.35	8042.21	103.14	0.0	conventional	2015	Albany
3	2015-12-06	1.08	78992.15	1132.00	71976.41	72.58	5811.16	5677.40	133.76	0.0	conventional	2015	Albany
4	2015-11-29	1.28	51039.60	941.48	43838.39	75.78	6183.95	5986.26	197.69	0.0	conventional	2015	Albany
...	...	...	...	...	...	...	...	...	...	...	...	...	...
18244	2018-02-04	1.63	17074.83	2046.96	1529.20	0.00	13498.67	13066.82	431.85	0.0	organic	2018	WestTexNewMexico
18245	2018-01-28	1.71	13888.04	1191.70	3431.50	0.00	9264.84	8940.04	324.80	0.0	organic	2018	WestTexNewMexico
18246	2018-01-21	1.87	13766.76	1191.92	2452.79	727.94	9394.11	9351.80	42.31	0.0	organic	2018	WestTexNewMexico
18247	2018-01-14	1.93	16205.22	1527.63	2981.04	727.01	10969.54	10919.54	50.00	0.0	organic	2018	WestTexNewMexico
18248	2018-01-07	1.62	17489.58	2894.77	2356.13	224.53	12014.15	11988.14	26.01	0.0	organic	2018	WestTexNewMexico

	date	average_price	total_volume	4046	4225	4770	total_bags	small_bags	large_bags	type	year	region
18243	2018-02-11	1.57	15986.17	1924.28	1368.32	0.00	12693.57	12437.35	256.22	organic	2018	WestTexNewMexico
18244	2018-02-04	1.63	17074.83	2046.96	1529.20	0.00	13498.67	13066.82	431.85	organic	2018	WestTexNewMexico
18245	2018-01-28	1.71	13888.04	1191.70	3431.50	0.00	9264.84	8940.04	324.80	organic	2018	WestTexNewMexico
18246	2018-01-21	1.87	13766.76	1191.92	2452.79	727.94	9394.11	9351.80	42.31	organic	2018	WestTexNewMexico
18247	2018-01-14	1.93	16205.22	1527.63	2981.04	727.01	10969.54	10919.54	50.00	organic	2018	WestTexNewMexico
18248	2018-01-07	1.62	17489.58	2894.77	2356.13	224.53	12014.15	11988.14	26.01	organic	2018	WestTexNewMexico

	date	average_price	total_volume	4046	4225	4770	total_bags	small_bags	large_bags	xlarge-bags	type	year	region
3322	2016-02-07	0.81	302337.87	4898.28	177380.94	30149.60	89909.05	22076.72	61180.85	6651.48	conventional	2016	CincinnatiDayton
12999	2016-06-26	1.72	3794.29	2437.58	118.08	25.30	1213.33	1213.33	0.00	0.00	organic	2016	Jacksonville
15456	2017-07-02	1.44	19735.70	435.47	3969.82	0.00	15330.41	343.28	14987.13	0.00	organic	2017	Detroit
4837	2016-12-18	1.15	93106.80	29593.90	26460.78	125.18	36926.94	35874.46	1052.48	0.00	conventional	2016	Roanoke
4395	2016-06-19	1.26	485945.76	53199.36	322695.72	27247.57	82803.11	80128.11	0.00	2675.00	conventional	2016	NorthernNewEngland
4679	2016-01-03	0.94	2171818.57	999388.89	840637.85	19331.34	312460.49	291734.88	19418.04	1307.57	conventional	2016	Plains
12011	2016-06-26	0.92	16159.03	2096.10	5307.76	0.00	8755.17	1075.61	7679.56	0.00	organic	2016	Atlanta

	count	mean	std	min	25%	50%	75%	max
average_price	18249.0	1.405978	4.026766e-01	0.44	1.10	1.37	1.66	3.25
total_volume	18249.0	850644.013009	3.453545e+06	84.56	10838.58	107376.76	432962.29	62505646.52
4046	18249.0	293008.424531	1.264989e+06	0.00	854.07	8645.30	111020.20	22743616.17
4225	18249.0	295154.568356	1.204120e+06	0.00	3008.78	29061.02	150206.86	20470572.61
4770	18249.0	22839.735993	1.074641e+05	0.00	0.00	184.99	6243.42	2546439.11
total_bags	18249.0	239639.202060	9.862424e+05	0.00	5088.64	39743.83	110783.37	19373134.37
small_bags	18249.0	182194.686696	7.461785e+05	0.00	2849.42	26362.82	83337.67	13384586.80
large_bags	18249.0	54338.088145	2.439660e+05	0.00	127.47	2647.71	22029.25	5719096.61
xlarge-bags	18249.0	3106.426507	1.769289e+04	0.00	0.00	0.00	132.50	551693.65
year	18249.0	2016.147899	9.399385e-01	2015.00	2015.00	2016.00	2017.00	2018.00

	total_volume	4046	4225	4770	total_bags	small_bags	large_bags
0	64,236.62	1,036.74	54,454.85	48.16	8,696.87	8,603.62	93.25
1	54,876.98	674.28	44,638.81	58.33	9,505.56	9,408.07	97.49
2	118,220.22	794.70	109,149.67	130.50	8,145.35	8,042.21	103.14
3	78,992.15	1,132.00	71,976.41	72.58	5,811.16	5,677.40	133.76
4	51,039.60	941.48	43,838.39	75.78	6,183.95	5,986.26	197.69

	average_price	total_volume	4046	4225	4770	total_bags	small_bags	large_bags	xlarge-bags	year
average_price	1.000000	-0.192752	-0.208317	-0.172928	-0.179446	-0.177088	-0.174730	-0.172940	-0.117592	0.093197
total_volume	-0.192752	1.000000	0.977863	0.974181	0.872202	0.963047	0.967238	0.880640	0.747157	0.017193
4046	-0.208317	0.977863	1.000000	0.926110	0.833389	0.920057	0.925280	0.838645	0.699377	0.003353
4225	-0.172928	0.974181	0.926110	1.000000	0.887855	0.905787	0.916031	0.810015	0.688809	-0.009559
4770	-0.179446	0.872202	0.833389	0.887855	1.000000	0.792314	0.802733	0.698471	0.679861	-0.036531
total_bags	-0.177088	0.963047	0.920057	0.905787	0.792314	1.000000	0.994335	0.943009	0.804233	0.071552
small_bags	-0.174730	0.967238	0.925280	0.916031	0.802733	0.994335	1.000000	0.902589	0.806845	0.063915
large_bags	-0.172940	0.880640	0.838645	0.810015	0.698471	0.943009	0.902589	1.000000	0.710858	0.087891
xlarge-bags	-0.117592	0.747157	0.699377	0.688809	0.679861	0.804233	0.806845	0.710858	1.000000	0.081033
year	0.093197	0.017193	0.003353	-0.009559	-0.036531	0.071552	0.063915	0.087891	0.081033	1.000000

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	...	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.7657	0.7652	0.7694	0.7663	0.1940	0.0766	0.1346	0.0986	0.0376	...	0.0911	0.313927	0.0098	0.1376	-12.2819	2.0139	2025-02-24-0715	2025-02-24-0716	0 day(s), 0 hour(s), 0 minute(s), 43 second(s)	0:00:43.667105
1	random_forest_regressor	0.7634	0.7629	0.7614	0.7637	0.1950	0.0774	0.1380	0.1014	0.0380	...	0.0971	0.313927	0.0076	0.1382	-12.2366	2.0144	2025-02-24-0719	2025-02-24-0722	0 day(s), 0 hour(s), 3 minute(s), 1 second(s)	0:03:01.143327
2	random_forest_regressor_2	0.7441	0.7435	0.7459	0.7443	0.2028	0.0804	0.1448	0.1065	0.0411	...	0.1048	0.313927	0.0058	0.1438	-12.1796	2.0190	2025-02-24-0722	2025-02-24-0725	0 day(s), 0 hour(s), 2 minute(s), 31 second(s)	0:02:31.139493
3	catboost_regressor	0.7264	0.7258	0.7345	0.7265	0.2097	0.0834	0.1529	0.1122	0.0440	...	0.1133	0.313927	0.0046	0.1487	-12.1564	2.0289	2025-02-24-0726	2025-02-24-0728	0 day(s), 0 hour(s), 1 minute(s), 26 second(s)	0:01:26.654634
4	xgb_regressor_2	0.7201	0.7195	0.7268	0.7202	0.2121	0.0842	0.1529	0.1121	0.0450	...	0.1102	0.313927	0.0045	0.1504	-12.2042	2.0315	2025-02-24-0719	2025-02-24-0719	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.642388
5	xgb_regressor_1	0.7201	0.7195	0.7268	0.7202	0.2121	0.0842	0.1529	0.1121	0.0450	...	0.1102	0.313927	0.0045	0.1504	-12.2042	2.0315	2025-02-24-0719	2025-02-24-0719	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.977416
6	lgbm_regressor	0.6933	0.6926	0.7024	0.6934	0.2220	0.0888	0.1640	0.1213	0.0493	...	0.1249	0.313927	0.0036	0.1574	-12.0810	2.0089	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.503374
7	hist_gradient_boosting_regressor	0.6894	0.6887	0.7006	0.6895	0.2234	0.0895	0.1652	0.1226	0.0499	...	0.1260	0.313927	0.0037	0.1584	-12.0840	2.0039	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.807350
8	knn_regressor_2	0.6587	0.6580	0.6722	0.6602	0.2342	0.0924	0.1617	0.1188	0.0548	...	0.1100	0.313927	0.0155	0.1660	-12.4276	2.0051	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.869274
9	knn_regressor_1	0.6587	0.6580	0.6722	0.6602	0.2342	0.0924	0.1617	0.1188	0.0548	...	0.1100	0.313927	0.0155	0.1660	-12.4276	2.0051	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.519692
10	gradient_boosting_regressor_1	0.5678	0.5668	0.5893	0.5678	0.2635	0.1060	0.2014	0.1503	0.0694	...	0.1609	0.313927	0.0021	0.1868	-11.9182	1.9978	2025-02-24-0716	2025-02-24-0716	0 day(s), 0 hour(s), 0 minute(s), 38 second(s)	0:00:38.181282
11	decision_tree_regressor	0.5627	0.5618	0.5194	0.5628	0.2651	0.1048	0.1739	0.1254	0.0703	...	0.1000	0.313927	0.0041	0.1879	-12.4771	2.0334	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.322157
12	decision_tree_pruned_regressor	0.4742	0.4730	0.4995	0.4742	0.2907	0.1171	0.2219	0.1659	0.0845	...	0.1802	0.313927	0.0032	0.2061	-11.9446	2.0000	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.904499
13	mlp_regressor_2	0.4630	0.4618	0.4768	0.4667	0.2937	0.1171	0.2212	0.1606	0.0863	...	0.1745	0.313927	-0.0244	0.2083	-11.4403	1.9916	2025-02-24-0725	2025-02-24-0726	0 day(s), 0 hour(s), 1 minute(s), 27 second(s)	0:01:27.593639
14	mlp_regressor_1	0.4380	0.4368	0.4638	0.4385	0.3005	0.1205	0.2301	0.1712	0.0903	...	0.1868	0.313927	0.0095	0.2130	-12.0364	2.0037	2025-02-24-0717	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 52 second(s)	0:00:52.771645
15	ada_boost_regressor	0.3543	0.3529	0.3905	0.4301	0.3221	0.1324	0.2627	0.2091	0.1037	...	0.2345	0.313927	0.1104	0.2284	-13.9920	1.7532	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 8 second(s)	0:00:08.258497
16	linear_regressor	0.0435	0.0414	0.0517	0.0436	0.3920	0.1618	0.3147	0.2464	0.1537	...	0.2767	0.313927	-0.0053	0.2779	-11.3516	2.0590	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.312107
17	bayesian_ridge_regressor	0.0433	0.0412	0.0515	0.0434	0.3921	0.1618	0.3148	0.2465	0.1537	...	0.2769	0.313927	-0.0053	0.2780	-11.3505	2.0597	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.210436
18	elasticnet_regressor_cv	0.0432	0.0411	0.0515	0.0433	0.3921	0.1619	0.3148	0.2465	0.1537	...	0.2768	0.313927	-0.0053	0.2780	-11.3511	2.0595	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.318008
19	ridge_regressor	0.0431	0.0410	0.0515	0.0433	0.3921	0.1619	0.3148	0.2465	0.1537	...	0.2766	0.313927	-0.0052	0.2780	-11.3516	2.0595	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.367521
20	ridge_regressor_cv	0.0431	0.0410	0.0515	0.0433	0.3921	0.1619	0.3148	0.2465	0.1537	...	0.2766	0.313927	-0.0052	0.2780	-11.3516	2.0595	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.801960
21	ridge_regressor_2	0.0431	0.0410	0.0515	0.0433	0.3921	0.1619	0.3148	0.2465	0.1537	...	0.2766	0.313927	-0.0052	0.2780	-11.3516	2.0595	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.175152
22	ard_regressor	0.0430	0.0409	0.0513	0.0431	0.3921	0.1619	0.3148	0.2466	0.1538	...	0.2768	0.313927	-0.0053	0.2780	-11.3503	2.0593	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.456839
23	huber_regressor	0.0316	0.0295	0.0434	0.0406	0.3944	0.1623	0.3139	0.2403	0.1556	...	0.2693	0.313927	-0.0380	0.2797	-10.7934	2.0372	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.051674
24	elasticnet_regressor_2	0.0272	0.0250	0.0278	0.0274	0.3954	0.1621	0.3169	0.2483	0.1563	...	0.2773	0.313927	-0.0055	0.2803	-11.2932	2.0799	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.235516
25	theil_sen_regressor	-61.3780	-61.5151	-50.1893	-59.6142	3.1658	0.2677	0.7912	0.7194	10.0222	...	0.2741	0.313927	-0.5323	2.2445	-66.4309	1.9551	2025-02-24-0718	2025-02-24-0719	0 day(s), 0 hour(s), 1 minute(s), 8 second(s)	0:01:08.474570
26	ransac_regressor	-12.7083	-12.7384	-19.9139	-12.5360	1.4841	0.2355	0.5286	0.4630	2.2025	...	0.2810	0.313927	-0.1664	1.0522	-22.5982	1.9510	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.882763
27	svm_regressor_linear_2	-0.2121	-0.2147	-0.1742	-0.0354	0.4413	0.1900	0.3597	0.3115	0.1947	...	0.3211	0.313927	0.1685	0.3129	-14.7146	1.7364	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.501946
28	svm_regressor_linear_1	-0.0093	-0.0115	0.0185	0.0032	0.4027	0.1652	0.3142	0.2409	0.1622	...	0.2650	0.313927	-0.0448	0.2855	-10.7659	2.0080	2025-02-24-0718	2025-02-24-0718	0 day(s), 0 hour(s), 0 minute(s), 7 second(s)	0:00:07.252729
29	gradient_boosting_regressor_2	-0.0002	-0.0024	0.0349	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0716	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 42 second(s)	0:00:42.733005
30	elastic_regressor_cv_2	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.085584
31	lasso_regressor_2	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.172075
32	elasticnet_regressor	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0715	2025-02-24-0715	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.191815
33	lasso_regressor_cv	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.739154
34	lasso_regressor	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	...	0.2849	0.313927	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0717	2025-02-24-0717	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.175969

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.6825	0.6814	0.6787	0.6842	0.2174	0.0828	0.1526	0.1031	0.0473	0.0068	0.3540	0.0991	0.1086	0.270	0.0158	0.1411	-15.9005	2.0047	2025-02-24-0730	2025-02-24-0731	0 day(s), 0 hour(s), 0 minute(s), 41 second(s)	0:00:41.630663
1	random_forest_regressor	0.6804	0.6793	0.6695	0.6816	0.2182	0.0828	0.1544	0.1038	0.0476	0.0069	0.3582	0.1002	0.1096	0.270	0.0135	0.1416	-15.8445	1.9985	2025-02-24-0734	2025-02-24-0735	0 day(s), 0 hour(s), 1 minute(s), 24 second(s)	0:01:24.015809
2	random_forest_regressor_2	0.6675	0.6663	0.6545	0.6684	0.2225	0.0845	0.1589	0.1069	0.0495	0.0071	0.3687	0.1032	0.1146	0.270	0.0118	0.1444	-15.7788	2.0012	2025-02-24-0735	2025-02-24-0736	0 day(s), 0 hour(s), 1 minute(s), 8 second(s)	0:01:08.108573
3	catboost_regressor	0.6565	0.6553	0.6506	0.6570	0.2262	0.0861	0.1657	0.1114	0.0512	0.0074	0.3846	0.1078	0.1251	0.270	0.0087	0.1468	-15.7192	2.0403	2025-02-24-0738	2025-02-24-0739	0 day(s), 0 hour(s), 1 minute(s), 11 second(s)	0:01:11.091304
4	xgb_regressor_1	0.6489	0.6477	0.6335	0.6493	0.2287	0.0872	0.1672	0.1122	0.0523	0.0076	0.3880	0.1089	0.1251	0.270	0.0080	0.1484	-15.7627	2.0179	2025-02-24-0733	2025-02-24-0733	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.147151
5	xgb_regressor_2	0.6489	0.6477	0.6335	0.6493	0.2287	0.0872	0.1672	0.1122	0.0523	0.0076	0.3880	0.1089	0.1251	0.270	0.0080	0.1484	-15.7627	2.0179	2025-02-24-0733	2025-02-24-0734	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.597624
6	hist_gradient_boosting_regressor	0.6350	0.6337	0.6289	0.6355	0.2331	0.0886	0.1716	0.1153	0.0544	0.0079	0.3982	0.1116	0.1279	0.270	0.0089	0.1513	-15.6890	2.0315	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.659597
7	lgbm_regressor	0.6336	0.6323	0.6292	0.6342	0.2336	0.0889	0.1727	0.1162	0.0546	0.0079	0.4009	0.1123	0.1288	0.270	0.0095	0.1516	-15.7009	2.0200	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.162337
8	knn_regressor_2	0.5755	0.5740	0.5485	0.5782	0.2514	0.0971	0.1786	0.1224	0.0632	0.0094	0.4143	0.1161	0.1280	0.270	0.0204	0.1632	-16.0213	2.0139	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.132037
9	knn_regressor_1	0.5755	0.5740	0.5485	0.5782	0.2514	0.0971	0.1786	0.1224	0.0632	0.0094	0.4143	0.1161	0.1280	0.270	0.0204	0.1632	-16.0213	2.0139	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.113118
10	gradient_boosting_regressor_1	0.4942	0.4925	0.5003	0.4948	0.2744	0.1043	0.2058	0.1391	0.0753	0.0109	0.4772	0.1337	0.1602	0.270	0.0098	0.1781	-15.5427	1.9918	2025-02-24-0731	2025-02-24-0731	0 day(s), 0 hour(s), 0 minute(s), 26 second(s)	0:00:26.472502
11	mlp_regressor_2	0.4346	0.4326	0.4388	0.4516	0.2902	0.1114	0.2218	0.1526	0.0842	0.0124	0.5144	0.1435	0.1751	0.270	0.0504	0.1883	-16.5257	1.9453	2025-02-24-0736	2025-02-24-0738	0 day(s), 0 hour(s), 1 minute(s), 27 second(s)	0:01:27.660473
12	mlp_regressor_1	0.4144	0.4123	0.4191	0.4144	0.2953	0.1119	0.2207	0.1482	0.0872	0.0125	0.5118	0.1427	0.1720	0.270	0.0013	0.1917	-15.3094	2.0007	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 30 second(s)	0:00:30.582822
13	decision_tree_pruned_regressor	0.3957	0.3936	0.3965	0.3964	0.3000	0.1141	0.2274	0.1529	0.0900	0.0130	0.5273	0.1469	0.1778	0.270	0.0100	0.1947	-15.5750	1.9705	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.704397
14	decision_tree_regressor	0.3486	0.3464	0.3101	0.3492	0.3114	0.1169	0.2091	0.1386	0.0970	0.0137	0.4851	0.1334	0.1300	0.270	0.0094	0.2021	-16.1247	1.9406	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.705608
15	huber_regressor	0.1860	0.1832	0.1896	0.1865	0.3481	0.1343	0.2696	0.1851	0.1212	0.0180	0.6251	0.1766	0.2226	0.270	-0.0086	0.2260	-14.9585	1.9748	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.604161
16	linear_regressor	0.1857	0.1829	0.1933	0.1870	0.3482	0.1349	0.2720	0.1893	0.1212	0.0182	0.6305	0.1780	0.2260	0.270	0.0137	0.2260	-15.4267	1.9718	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.148288
17	elasticnet_regressor_cv	0.1854	0.1826	0.1931	0.1867	0.3483	0.1349	0.2721	0.1894	0.1213	0.0182	0.6307	0.1780	0.2255	0.270	0.0138	0.2261	-15.4263	1.9717	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.135980
18	bayesian_ridge_regressor	0.1853	0.1825	0.1931	0.1866	0.3483	0.1349	0.2721	0.1894	0.1213	0.0182	0.6308	0.1780	0.2257	0.270	0.0138	0.2261	-15.4273	1.9716	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.167477
19	ard_regressor	0.1853	0.1826	0.1931	0.1866	0.3483	0.1349	0.2721	0.1894	0.1213	0.0182	0.6308	0.1780	0.2257	0.270	0.0138	0.2261	-15.4269	1.9716	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.263121
20	ridge_regressor_cv	0.1853	0.1825	0.1931	0.1866	0.3483	0.1350	0.2721	0.1894	0.1213	0.0182	0.6308	0.1780	0.2259	0.270	0.0138	0.2261	-15.4283	1.9716	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.895669
21	ridge_regressor_2	0.1853	0.1825	0.1931	0.1866	0.3483	0.1350	0.2721	0.1894	0.1213	0.0182	0.6308	0.1780	0.2259	0.270	0.0138	0.2261	-15.4283	1.9716	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.143200
22	ridge_regressor	0.1853	0.1825	0.1931	0.1866	0.3483	0.1350	0.2721	0.1894	0.1213	0.0182	0.6308	0.1780	0.2259	0.270	0.0138	0.2261	-15.4283	1.9716	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.250136
23	svm_regressor_linear_1	0.1843	0.1816	0.1845	0.1863	0.3485	0.1342	0.2692	0.1839	0.1215	0.0180	0.6240	0.1762	0.2191	0.270	-0.0170	0.2262	-14.7741	1.9705	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.198836
24	ada_boost_regressor	0.1822	0.1794	0.2334	0.3310	0.3490	0.1360	0.2865	0.2074	0.1218	0.0185	0.6645	0.1841	0.2631	0.270	0.1489	0.2265	-18.5754	1.6120	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 7 second(s)	0:00:07.384942
25	elasticnet_regressor_2	0.1579	0.1550	0.1591	0.1591	0.3541	0.1369	0.2762	0.1930	0.1254	0.0187	0.6402	0.1801	0.2259	0.270	0.0130	0.2298	-15.2866	1.9848	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.288644
26	lasso_regressor_2	0.0945	0.0914	0.0935	0.0956	0.3672	0.1425	0.2879	0.2028	0.1348	0.0203	0.6675	0.1883	0.2426	0.270	0.0128	0.2383	-15.2256	1.9989	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.152300
27	theil_sen_regressor	-0.9062	-0.9127	-0.8236	-0.9062	0.5328	0.2268	0.3840	0.2787	0.2838	0.0514	0.8906	0.2831	0.2748	0.270	0.0004	0.3458	-16.3702	2.0001	2025-02-24-0732	2025-02-24-0733	0 day(s), 0 hour(s), 0 minute(s), 56 second(s)	0:00:56.871327
28	ransac_regressor	-0.1602	-0.1642	-0.5250	-0.0842	0.4156	0.1748	0.3128	0.2147	0.1728	0.0305	0.7251	0.2262	0.2416	0.270	-0.1063	0.2698	-13.4681	1.8209	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.932832
29	svm_regressor_linear_2	-0.0601	-0.0638	-0.1065	0.0733	0.3973	0.1612	0.2994	0.1946	0.1579	0.0260	0.6940	0.2083	0.2300	0.270	-0.1409	0.2579	-12.5847	1.6855	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.380196
30	lasso_regressor_cv	-0.0011	-0.0045	-0.0019	-0.0000	0.3861	0.1503	0.3044	0.2154	0.1491	0.0226	0.7056	0.1989	0.2632	0.270	0.0126	0.2506	-15.2025	2.0142	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.636627
31	gradient_boosting_regressor_2	-0.0011	-0.0045	0.0515	-0.0000	0.3861	0.1503	0.3044	0.2154	0.1491	0.0226	0.7056	0.1989	0.2632	0.270	0.0126	0.2506	-15.2025	2.0142	2025-02-24-0731	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 31 second(s)	0:00:31.178545
32	elastic_regressor_cv_2	-0.0011	-0.0045	-0.0019	-0.0000	0.3861	0.1503	0.3044	0.2154	0.1491	0.0226	0.7056	0.1989	0.2632	0.270	0.0126	0.2506	-15.2025	2.0142	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.007732
33	elasticnet_regressor	-0.0011	-0.0045	-0.0019	-0.0000	0.3861	0.1503	0.3044	0.2154	0.1491	0.0226	0.7056	0.1989	0.2632	0.270	0.0126	0.2506	-15.2025	2.0142	2025-02-24-0730	2025-02-24-0730	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.185335
34	lasso_regressor	-0.0011	-0.0045	-0.0019	-0.0000	0.3861	0.1503	0.3044	0.2154	0.1491	0.0226	0.7056	0.1989	0.2632	0.270	0.0126	0.2506	-15.2025	2.0142	2025-02-24-0732	2025-02-24-0732	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.180616

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.7732	0.7727	0.7618	0.7735	0.1878	0.0751	0.1306	0.0969	0.0353	0.0056	0.2956	0.0933	0.0882	0.296	0.0065	0.1340	-12.4620	2.0112	2025-02-24-0739	2025-02-24-0740	0 day(s), 0 hour(s), 1 minute(s), 0 second(s)	0:01:00.843443
1	random_forest_regressor	0.7641	0.7636	0.7536	0.7643	0.1915	0.0765	0.1349	0.1002	0.0367	0.0059	0.3053	0.0966	0.0944	0.296	0.0055	0.1366	-12.4319	1.9909	2025-02-24-0745	2025-02-24-0747	0 day(s), 0 hour(s), 1 minute(s), 59 second(s)	0:01:59.930896
2	random_forest_regressor_2	0.7503	0.7498	0.7400	0.7504	0.1971	0.0789	0.1408	0.1049	0.0388	0.0062	0.3186	0.1010	0.1001	0.296	0.0039	0.1406	-12.3764	1.9870	2025-02-24-0747	2025-02-24-0748	0 day(s), 0 hour(s), 1 minute(s), 23 second(s)	0:01:23.106899
3	catboost_regressor	0.7300	0.7293	0.7278	0.7300	0.2049	0.0823	0.1500	0.1117	0.0420	0.0068	0.3395	0.1079	0.1107	0.296	0.0032	0.1462	-12.3789	1.9847	2025-02-24-0749	2025-02-24-0750	0 day(s), 0 hour(s), 1 minute(s), 0 second(s)	0:01:00.830499
4	xgb_regressor_2	0.7296	0.7290	0.7188	0.7296	0.2051	0.0823	0.1484	0.1098	0.0421	0.0068	0.3359	0.1066	0.1071	0.296	-0.0010	0.1463	-12.3410	1.9808	2025-02-24-0745	2025-02-24-0745	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.138823
5	xgb_regressor_1	0.7296	0.7290	0.7188	0.7296	0.2051	0.0823	0.1484	0.1098	0.0421	0.0068	0.3359	0.1066	0.1071	0.296	-0.0010	0.1463	-12.3410	1.9808	2025-02-24-0745	2025-02-24-0745	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.722361
6	lgbm_regressor	0.6989	0.6982	0.6957	0.6990	0.2164	0.0875	0.1610	0.1209	0.0468	0.0076	0.3643	0.1164	0.1244	0.296	0.0025	0.1544	-12.3047	1.9968	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.603814
7	hist_gradient_boosting_regressor	0.6945	0.6938	0.6936	0.6946	0.2180	0.0880	0.1622	0.1216	0.0475	0.0077	0.3672	0.1170	0.1247	0.296	0.0036	0.1555	-12.3231	1.9951	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 5 second(s)	0:00:05.758350
8	knn_regressor_2	0.6638	0.6631	0.6533	0.6646	0.2287	0.0926	0.1612	0.1212	0.0523	0.0086	0.3648	0.1149	0.1080	0.296	0.0112	0.1631	-12.5589	2.0104	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.660658
9	knn_regressor_1	0.6638	0.6631	0.6533	0.6646	0.2287	0.0926	0.1612	0.1212	0.0523	0.0086	0.3648	0.1149	0.1080	0.296	0.0112	0.1631	-12.5589	2.0104	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.466490
10	gradient_boosting_regressor_1	0.5784	0.5774	0.5771	0.5786	0.2561	0.1040	0.1963	0.1489	0.0656	0.0108	0.4442	0.1420	0.1581	0.296	0.0056	0.1827	-12.2377	1.9874	2025-02-24-0740	2025-02-24-0741	0 day(s), 0 hour(s), 0 minute(s), 43 second(s)	0:00:43.085310
11	decision_tree_regressor	0.4894	0.4882	0.4838	0.4895	0.2818	0.1120	0.1859	0.1360	0.0794	0.0125	0.4208	0.1312	0.1100	0.296	0.0050	0.2010	-12.7135	1.9632	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.475351
12	mlp_regressor_2	0.4890	0.4878	0.4775	0.4896	0.2819	0.1131	0.2137	0.1584	0.0795	0.0128	0.4835	0.1522	0.1710	0.296	-0.0098	0.2011	-11.9331	2.0256	2025-02-24-0748	2025-02-24-0749	0 day(s), 0 hour(s), 1 minute(s), 5 second(s)	0:01:05.315619
13	decision_tree_pruned_regressor	0.4832	0.4821	0.4874	0.4836	0.2835	0.1149	0.2172	0.1641	0.0804	0.0132	0.4914	0.1563	0.1784	0.296	0.0071	0.2022	-12.2845	1.9992	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.897245
14	mlp_regressor_1	0.4720	0.4709	0.4355	0.4721	0.2866	0.1157	0.2194	0.1646	0.0821	0.0134	0.4966	0.1573	0.1765	0.296	-0.0012	0.2044	-12.1391	2.0209	2025-02-24-0742	2025-02-24-0743	0 day(s), 0 hour(s), 0 minute(s), 59 second(s)	0:00:59.210894
15	ada_boost_regressor	0.3267	0.3252	0.3420	0.4362	0.3236	0.1335	0.2651	0.2144	0.1047	0.0178	0.5999	0.1890	0.2332	0.296	0.1305	0.2308	-14.7104	1.6854	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 14 second(s)	0:00:14.767303
16	huber_regressor	0.1251	0.1232	0.1178	0.1258	0.3689	0.1522	0.2957	0.2325	0.1361	0.0232	0.6694	0.2162	0.2606	0.296	-0.0102	0.2631	-11.5603	1.9789	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.707036
17	linear_regressor	0.1248	0.1229	0.1239	0.1268	0.3689	0.1536	0.2974	0.2386	0.1361	0.0236	0.6730	0.2181	0.2650	0.296	0.0174	0.2632	-12.0802	1.9752	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.462134
18	ridge_regressor_cv	0.1243	0.1223	0.1236	0.1262	0.3690	0.1536	0.2975	0.2387	0.1362	0.0236	0.6733	0.2181	0.2657	0.296	0.0175	0.2633	-12.0815	1.9742	2025-02-24-0744	2025-02-24-0744	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.847924
19	ridge_regressor	0.1243	0.1223	0.1236	0.1262	0.3690	0.1536	0.2975	0.2387	0.1362	0.0236	0.6733	0.2181	0.2657	0.296	0.0175	0.2633	-12.0815	1.9742	2025-02-24-0743	2025-02-24-0744	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.336141
20	ridge_regressor_2	0.1243	0.1223	0.1236	0.1262	0.3690	0.1536	0.2975	0.2387	0.1362	0.0236	0.6733	0.2181	0.2657	0.296	0.0175	0.2633	-12.0815	1.9742	2025-02-24-0744	2025-02-24-0744	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.111903
21	bayesian_ridge_regressor	0.1242	0.1222	0.1236	0.1261	0.3691	0.1536	0.2975	0.2387	0.1362	0.0236	0.6733	0.2181	0.2654	0.296	0.0175	0.2633	-12.0808	1.9743	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.537284
22	elasticnet_regressor_cv	0.1241	0.1222	0.1237	0.1261	0.3691	0.1536	0.2975	0.2387	0.1362	0.0236	0.6733	0.2181	0.2656	0.296	0.0174	0.2633	-12.0798	1.9743	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.837246
23	ard_regressor	0.1236	0.1216	0.1231	0.1255	0.3692	0.1536	0.2976	0.2386	0.1363	0.0236	0.6735	0.2181	0.2641	0.296	0.0173	0.2634	-12.0783	1.9740	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.206992
24	svm_regressor_linear_1	0.1176	0.1156	0.1104	0.1179	0.3705	0.1523	0.2960	0.2328	0.1372	0.0232	0.6700	0.2147	0.2579	0.296	-0.0069	0.2643	-11.5819	1.9829	2025-02-24-0744	2025-02-24-0744	0 day(s), 0 hour(s), 0 minute(s), 6 second(s)	0:00:06.372905
25	elasticnet_regressor_2	0.0982	0.0962	0.0985	0.1001	0.3745	0.1548	0.2997	0.2398	0.1402	0.0240	0.6782	0.2167	0.2632	0.296	0.0169	0.2671	-11.9871	1.9816	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.169178
26	lasso_regressor_2	0.0482	0.0461	0.0486	0.0501	0.3847	0.1594	0.3090	0.2484	0.1480	0.0254	0.6993	0.2237	0.2721	0.296	0.0170	0.2744	-11.9514	1.9898	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.188611
27	theil_sen_regressor	-5.1155	-5.1293	-5.1936	-4.7703	0.9753	0.2563	0.4789	0.4140	0.9511	0.0657	1.0837	0.3605	0.2663	0.296	-0.2317	0.6957	-14.2850	1.8721	2025-02-24-0744	2025-02-24-0745	0 day(s), 0 hour(s), 0 minute(s), 57 second(s)	0:00:57.934997
28	ransac_regressor	-2.4463	-2.4541	-1.0613	-2.2851	0.7321	0.2370	0.4122	0.3415	0.5360	0.0562	0.9331	0.3206	0.2636	0.296	-0.1584	0.5222	-12.3665	1.8995	2025-02-24-0743	2025-02-24-0743	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.976502
29	svm_regressor_linear_2	-0.2002	-0.2029	-0.1269	-0.0920	0.4320	0.1947	0.3519	0.3038	0.1867	0.0379	0.7964	0.2764	0.3095	0.296	0.1297	0.3082	-14.6807	1.7921	2025-02-24-0744	2025-02-24-0744	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.183159
30	gradient_boosting_regressor_2	-0.0019	-0.0041	0.1770	0.0000	0.3947	0.1637	0.3178	0.2562	0.1558	0.0268	0.7193	0.2299	0.2810	0.296	0.0171	0.2816	-11.9460	1.9955	2025-02-24-0741	2025-02-24-0742	0 day(s), 0 hour(s), 1 minute(s), 2 second(s)	0:01:02.399685
31	lasso_regressor_cv	-0.0019	-0.0041	-0.0007	0.0000	0.3947	0.1637	0.3178	0.2562	0.1558	0.0268	0.7193	0.2299	0.2810	0.296	0.0171	0.2816	-11.9460	1.9955	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.988236
32	elastic_regressor_cv_2	-0.0019	-0.0041	-0.0007	0.0000	0.3947	0.1637	0.3178	0.2562	0.1558	0.0268	0.7193	0.2299	0.2810	0.296	0.0171	0.2816	-11.9460	1.9955	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.672039
33	elasticnet_regressor	-0.0019	-0.0041	-0.0007	0.0000	0.3947	0.1637	0.3178	0.2562	0.1558	0.0268	0.7193	0.2299	0.2810	0.296	0.0171	0.2816	-11.9460	1.9955	2025-02-24-0739	2025-02-24-0739	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.315093
34	lasso_regressor	-0.0019	-0.0041	-0.0007	0.0000	0.3947	0.1637	0.3178	0.2562	0.1558	0.0268	0.7193	0.2299	0.2810	0.296	0.0171	0.2816	-11.9460	1.9955	2025-02-24-0742	2025-02-24-0742	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.292210

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.5823	0.5804	0.5754	0.5825	0.2402	0.0880	0.1677	0.1037	0.0577	0.0078	0.4083	0.1020	0.1125	0.275	-0.0054	0.1454	-19.1720	1.9675	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 19 second(s)	0:00:19.275262
1	random_forest_regressor	0.5751	0.5731	0.5710	0.5754	0.2423	0.0888	0.1695	0.1047	0.0587	0.0079	0.4125	0.1029	0.1213	0.275	-0.0063	0.1466	-19.1300	1.9602	2025-02-24-0753	2025-02-24-0754	0 day(s), 0 hour(s), 0 minute(s), 57 second(s)	0:00:57.334004
2	catboost_regressor	0.5739	0.5718	0.5612	0.5745	0.2426	0.0887	0.1738	0.1071	0.0589	0.0079	0.4231	0.1054	0.1281	0.275	-0.0091	0.1468	-19.0499	1.9977	2025-02-24-0755	2025-02-24-0756	0 day(s), 0 hour(s), 1 minute(s), 3 second(s)	0:01:03.659437
3	random_forest_regressor_2	0.5638	0.5618	0.5556	0.5641	0.2455	0.0900	0.1751	0.1086	0.0603	0.0081	0.4263	0.1066	0.1271	0.275	-0.0057	0.1486	-19.1101	1.9620	2025-02-24-0754	2025-02-24-0755	0 day(s), 0 hour(s), 0 minute(s), 41 second(s)	0:00:41.016791
4	hist_gradient_boosting_regressor	0.5527	0.5506	0.5394	0.5532	0.2486	0.0912	0.1819	0.1128	0.0618	0.0083	0.4429	0.1108	0.1371	0.275	-0.0083	0.1504	-19.0324	1.9721	2025-02-24-0751	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.035836
5	lgbm_regressor	0.5489	0.5468	0.5408	0.5498	0.2496	0.0917	0.1821	0.1129	0.0623	0.0084	0.4433	0.1110	0.1387	0.275	-0.0107	0.1511	-18.9804	1.9885	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.839212
6	xgb_regressor_2	0.5412	0.5391	0.5324	0.5419	0.2518	0.0923	0.1798	0.1109	0.0634	0.0085	0.4377	0.1092	0.1339	0.275	-0.0096	0.1524	-19.1077	1.9635	2025-02-24-0753	2025-02-24-0753	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.648352
7	xgb_regressor_1	0.5412	0.5391	0.5324	0.5419	0.2518	0.0923	0.1798	0.1109	0.0634	0.0085	0.4377	0.1092	0.1339	0.275	-0.0096	0.1524	-19.1077	1.9635	2025-02-24-0753	2025-02-24-0753	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.443644
8	knn_regressor_2	0.4798	0.4773	0.4693	0.4804	0.2681	0.0987	0.1866	0.1156	0.0719	0.0098	0.4544	0.1133	0.1300	0.275	-0.0087	0.1622	-19.1653	1.9726	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.615439
9	knn_regressor_1	0.4798	0.4773	0.4693	0.4804	0.2681	0.0987	0.1866	0.1156	0.0719	0.0098	0.4544	0.1133	0.1300	0.275	-0.0087	0.1622	-19.1653	1.9726	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.611409
10	gradient_boosting_regressor_1	0.4001	0.3972	0.4072	0.4010	0.2879	0.1057	0.2151	0.1341	0.0829	0.0112	0.5237	0.1309	0.1662	0.275	-0.0112	0.1742	-18.8066	1.9705	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 15 second(s)	0:00:15.879554
11	mlp_regressor_1	0.3298	0.3266	0.3341	0.3300	0.3043	0.1119	0.2321	0.1453	0.0926	0.0125	0.5651	0.1413	0.1900	0.275	-0.0052	0.1842	-18.9077	1.9959	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 21 second(s)	0:00:21.797133
12	mlp_regressor_2	0.3231	0.3199	0.3565	0.3240	0.3058	0.1129	0.2324	0.1470	0.0935	0.0128	0.5657	0.1416	0.1913	0.275	0.0110	0.1851	-19.3009	2.0027	2025-02-24-0755	2025-02-24-0755	0 day(s), 0 hour(s), 0 minute(s), 49 second(s)	0:00:49.857986
13	decision_tree_pruned_regressor	0.2525	0.2490	0.2774	0.2536	0.3214	0.1186	0.2446	0.1529	0.1033	0.0141	0.5955	0.1488	0.1992	0.275	-0.0118	0.1945	-18.7621	1.9708	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.402070
14	decision_tree_regressor	0.1964	0.1926	0.1555	0.1966	0.3332	0.1230	0.2275	0.1419	0.1110	0.0151	0.5540	0.1371	0.1500	0.275	0.0048	0.2016	-19.8295	2.0123	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.946191
15	ada_boost_regressor	0.1550	0.1510	0.0864	0.2251	0.3417	0.1296	0.2774	0.1865	0.1167	0.0168	0.6751	0.1690	0.2473	0.275	0.0984	0.2068	-21.3619	1.8070	2025-02-24-0750	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.324284
16	gradient_boosting_regressor_2	0.1156	0.1114	0.0688	0.1161	0.3496	0.1305	0.2723	0.1752	0.1222	0.0170	0.6627	0.1664	0.2285	0.275	-0.0088	0.2115	-18.6170	1.9857	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 17 second(s)	0:00:17.681063
17	ard_regressor	0.0933	0.0890	0.1031	0.0942	0.3539	0.1317	0.2763	0.1760	0.1253	0.0173	0.6727	0.1688	0.2244	0.275	-0.0111	0.2142	-18.6015	1.9739	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.306224
18	ridge_regressor_2	0.0927	0.0884	0.1030	0.0936	0.3541	0.1318	0.2764	0.1761	0.1254	0.0174	0.6731	0.1689	0.2251	0.275	-0.0112	0.2143	-18.5992	1.9733	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.134319
19	ridge_regressor	0.0927	0.0884	0.1030	0.0936	0.3541	0.1318	0.2764	0.1761	0.1254	0.0174	0.6731	0.1689	0.2251	0.275	-0.0112	0.2143	-18.5992	1.9733	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.225601
20	bayesian_ridge_regressor	0.0927	0.0884	0.1029	0.0936	0.3540	0.1318	0.2764	0.1761	0.1254	0.0174	0.6730	0.1689	0.2253	0.275	-0.0112	0.2143	-18.5985	1.9732	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.195386
21	elasticnet_regressor_cv	0.0927	0.0884	0.1029	0.0936	0.3540	0.1318	0.2764	0.1761	0.1253	0.0174	0.6731	0.1689	0.2250	0.275	-0.0112	0.2143	-18.5992	1.9733	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.176405
22	ridge_regressor_cv	0.0927	0.0884	0.1030	0.0936	0.3541	0.1318	0.2764	0.1761	0.1254	0.0174	0.6731	0.1689	0.2251	0.275	-0.0112	0.2143	-18.5992	1.9733	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.892843
23	linear_regressor	0.0924	0.0881	0.1029	0.0933	0.3541	0.1318	0.2764	0.1760	0.1254	0.0174	0.6730	0.1689	0.2265	0.275	-0.0113	0.2143	-18.5973	1.9711	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.134725
24	huber_regressor	0.0831	0.0788	0.0971	0.0935	0.3559	0.1318	0.2752	0.1724	0.1267	0.0174	0.6702	0.1681	0.2191	0.275	-0.0379	0.2154	-17.9655	1.9554	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.576220
25	svm_regressor_linear_1	0.0786	0.0743	0.0907	0.0924	0.3568	0.1319	0.2750	0.1718	0.1273	0.0174	0.6695	0.1678	0.2196	0.275	-0.0436	0.2159	-17.8166	1.9505	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.413535
26	elasticnet_regressor_2	0.0587	0.0543	0.0584	0.0591	0.3606	0.1346	0.2822	0.1817	0.1300	0.0181	0.6868	0.1722	0.2332	0.275	-0.0070	0.2182	-18.6126	1.9778	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.231186
27	ransac_regressor	-0.9637	-0.9730	-6103187298813.8760	-0.9631	0.5209	0.1929	0.3660	0.2364	0.2713	0.0372	0.8901	0.2209	0.2497	0.275	-0.0091	0.3152	-19.5484	2.0014	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.780732
28	theil_sen_regressor	-0.1934	-0.1991	-0.2225	-0.1934	0.4061	0.1569	0.3102	0.2000	0.1649	0.0246	0.7557	0.1950	0.2453	0.275	0.0025	0.2457	-19.3136	1.9767	2025-02-24-0752	2025-02-24-0753	0 day(s), 0 hour(s), 0 minute(s), 47 second(s)	0:00:47.795201
29	svm_regressor_linear_2	-0.1521	-0.1576	-0.2168	-0.0820	0.3990	0.1490	0.3097	0.1899	0.1592	0.0222	0.7542	0.1912	0.2478	0.275	-0.0984	0.2414	-16.6736	1.8445	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.336700
30	elastic_regressor_cv_2	-0.0003	-0.0050	-0.0016	0.0000	0.3717	0.1391	0.2906	0.1883	0.1382	0.0194	0.7073	0.1773	0.2362	0.275	-0.0062	0.2250	-18.6153	1.9811	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.704121
31	lasso_regressor_2	-0.0003	-0.0050	-0.0016	0.0000	0.3717	0.1391	0.2906	0.1883	0.1382	0.0194	0.7073	0.1773	0.2362	0.275	-0.0062	0.2250	-18.6153	1.9811	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.134786
32	elasticnet_regressor	-0.0003	-0.0050	-0.0016	0.0000	0.3717	0.1391	0.2906	0.1883	0.1382	0.0194	0.7073	0.1773	0.2362	0.275	-0.0062	0.2250	-18.6153	1.9811	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.136933
33	lasso_regressor_cv	-0.0003	-0.0050	-0.0016	0.0000	0.3717	0.1391	0.2906	0.1883	0.1382	0.0194	0.7073	0.1773	0.2362	0.275	-0.0062	0.2250	-18.6153	1.9811	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.599016
34	lasso_regressor	-0.0003	-0.0050	-0.0016	0.0000	0.3717	0.1391	0.2906	0.1883	0.1382	0.0194	0.7073	0.1773	0.2362	0.275	-0.0062	0.2250	-18.6153	1.9811	2025-02-24-0752	2025-02-24-0752	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.206755

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.7288	0.7280	0.7383	0.7288	0.2138	0.0820	0.1463	0.1015	0.0457	0.0067	0.3242	0.0986	0.0980	0.300	0.0015	0.1445	-12.7016	2.0774	2025-02-24-0757	2025-02-24-0758	0 day(s), 0 hour(s), 0 minute(s), 45 second(s)	0:00:45.592145
1	random_forest_regressor	0.7191	0.7183	0.7302	0.7191	0.2175	0.0840	0.1494	0.1041	0.0473	0.0071	0.3311	0.1014	0.1025	0.300	-0.0005	0.1470	-12.6674	2.0763	2025-02-24-0802	2025-02-24-0803	0 day(s), 0 hour(s), 1 minute(s), 28 second(s)	0:01:28.817383
2	random_forest_regressor_2	0.7083	0.7075	0.7156	0.7084	0.2217	0.0856	0.1545	0.1078	0.0491	0.0073	0.3425	0.1050	0.1087	0.300	-0.0026	0.1498	-12.5971	2.0642	2025-02-24-0803	2025-02-24-0804	0 day(s), 0 hour(s), 1 minute(s), 1 second(s)	0:01:01.072849
3	catboost_regressor	0.6928	0.6919	0.7102	0.6929	0.2275	0.0876	0.1631	0.1132	0.0518	0.0077	0.3614	0.1105	0.1202	0.300	-0.0044	0.1538	-12.5755	2.0329	2025-02-24-0805	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 56 second(s)	0:00:56.314686
4	xgb_regressor_2	0.6817	0.6807	0.6983	0.6817	0.2316	0.0897	0.1660	0.1153	0.0536	0.0081	0.3679	0.1126	0.1213	0.300	-0.0029	0.1565	-12.6503	2.0182	2025-02-24-0801	2025-02-24-0802	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.546742
5	xgb_regressor_1	0.6817	0.6807	0.6983	0.6817	0.2316	0.0897	0.1660	0.1153	0.0536	0.0081	0.3679	0.1126	0.1213	0.300	-0.0029	0.1565	-12.6503	2.0182	2025-02-24-0801	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.759707
6	lgbm_regressor	0.6758	0.6748	0.6833	0.6760	0.2337	0.0904	0.1697	0.1188	0.0546	0.0082	0.3761	0.1156	0.1278	0.300	-0.0060	0.1580	-12.5035	2.0329	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.842404
7	hist_gradient_boosting_regressor	0.6728	0.6718	0.6807	0.6729	0.2348	0.0909	0.1707	0.1195	0.0551	0.0083	0.3782	0.1164	0.1301	0.300	-0.0042	0.1587	-12.5293	2.0338	2025-02-24-0759	2025-02-24-0759	0 day(s), 0 hour(s), 0 minute(s), 5 second(s)	0:00:05.182555
8	knn_regressor_2	0.6327	0.6317	0.6166	0.6327	0.2487	0.0964	0.1738	0.1209	0.0619	0.0093	0.3853	0.1170	0.1160	0.300	0.0011	0.1681	-12.7167	2.1140	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.412212
9	knn_regressor_1	0.6327	0.6317	0.6166	0.6327	0.2487	0.0964	0.1738	0.1209	0.0619	0.0093	0.3853	0.1170	0.1160	0.300	0.0011	0.1681	-12.7167	2.1140	2025-02-24-0759	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.933448
10	gradient_boosting_regressor_1	0.5765	0.5753	0.5687	0.5767	0.2671	0.1033	0.1991	0.1399	0.0713	0.0107	0.4412	0.1352	0.1555	0.300	-0.0063	0.1806	-12.3547	2.0026	2025-02-24-0758	2025-02-24-0759	0 day(s), 0 hour(s), 0 minute(s), 36 second(s)	0:00:36.228211
11	decision_tree_pruned_regressor	0.4950	0.4935	0.4777	0.4951	0.2917	0.1138	0.2201	0.1560	0.0851	0.0129	0.4877	0.1497	0.1748	0.300	-0.0046	0.1972	-12.3905	1.9992	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.783536
12	mlp_regressor_1	0.4795	0.4780	0.4310	0.4829	0.2961	0.1137	0.2212	0.1527	0.0877	0.0129	0.4900	0.1488	0.1746	0.300	-0.0238	0.2002	-11.9786	1.9913	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 47 second(s)	0:00:47.852844
13	decision_tree_regressor	0.4444	0.4428	0.4660	0.4445	0.3059	0.1165	0.1968	0.1341	0.0936	0.0136	0.4361	0.1300	0.1100	0.300	0.0032	0.2068	-13.0741	2.0247	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.022047
14	mlp_regressor_2	0.4411	0.4395	0.4615	0.4417	0.3068	0.1245	0.2334	0.1667	0.0942	0.0155	0.5169	0.1654	0.1875	0.300	-0.0097	0.2074	-12.4394	2.0085	2025-02-24-0804	2025-02-24-0805	0 day(s), 0 hour(s), 0 minute(s), 57 second(s)	0:00:57.970660
15	ada_boost_regressor	0.4012	0.3995	0.3527	0.4575	0.3176	0.1269	0.2559	0.1935	0.1009	0.0161	0.5669	0.1746	0.2276	0.300	0.0974	0.2147	-14.2414	1.7962	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 6 second(s)	0:00:06.150319
16	gradient_boosting_regressor_2	0.3113	0.3093	0.1189	0.3116	0.3406	0.1362	0.2671	0.1971	0.1160	0.0185	0.5917	0.1851	0.2238	0.300	-0.0069	0.2302	-12.1006	1.9530	2025-02-24-0759	2025-02-24-0759	0 day(s), 0 hour(s), 0 minute(s), 49 second(s)	0:00:49.643947
17	linear_regressor	0.1482	0.1458	0.1283	0.1483	0.3788	0.1542	0.3013	0.2267	0.1435	0.0238	0.6676	0.2129	0.2597	0.300	-0.0040	0.2561	-12.0541	1.9575	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.118839
18	bayesian_ridge_regressor	0.1466	0.1441	0.1300	0.1467	0.3792	0.1544	0.3016	0.2269	0.1438	0.0238	0.6683	0.2131	0.2591	0.300	-0.0043	0.2563	-12.0470	1.9539	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.211515
19	ridge_regressor_cv	0.1466	0.1442	0.1299	0.1467	0.3792	0.1544	0.3015	0.2269	0.1438	0.0238	0.6682	0.2132	0.2594	0.300	-0.0043	0.2563	-12.0478	1.9541	2025-02-24-0800	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.926338
20	ridge_regressor	0.1466	0.1442	0.1299	0.1467	0.3792	0.1544	0.3015	0.2269	0.1438	0.0238	0.6682	0.2132	0.2594	0.300	-0.0043	0.2563	-12.0479	1.9541	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.270296
21	ridge_regressor_2	0.1466	0.1442	0.1299	0.1467	0.3792	0.1544	0.3015	0.2269	0.1438	0.0238	0.6682	0.2132	0.2594	0.300	-0.0043	0.2563	-12.0479	1.9541	2025-02-24-0801	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.160086
22	ard_regressor	0.1465	0.1440	0.1297	0.1466	0.3792	0.1544	0.3015	0.2269	0.1438	0.0238	0.6682	0.2131	0.2590	0.300	-0.0043	0.2563	-12.0477	1.9550	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.326876
23	elasticnet_regressor_cv	0.1462	0.1437	0.1300	0.1463	0.3793	0.1543	0.3017	0.2270	0.1438	0.0238	0.6685	0.2131	0.2594	0.300	-0.0043	0.2564	-12.0433	1.9538	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.923927
24	elasticnet_regressor_2	0.1095	0.1069	0.1009	0.1096	0.3873	0.1550	0.3074	0.2307	0.1500	0.0240	0.6813	0.2118	0.2678	0.300	-0.0042	0.2618	-11.9609	1.9442	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.184988
25	huber_regressor	0.1015	0.0989	0.0871	0.1109	0.3891	0.1571	0.2997	0.2222	0.1514	0.0247	0.6642	0.2156	0.2522	0.300	-0.0398	0.2630	-11.5846	1.9425	2025-02-24-0759	2025-02-24-0759	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.614732
26	lasso_regressor_2	0.0484	0.0457	0.0438	0.0485	0.4004	0.1608	0.3188	0.2410	0.1603	0.0259	0.7065	0.2195	0.2761	0.300	-0.0034	0.2706	-11.9391	1.9366	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.159460
27	svm_regressor_linear_2	0.0326	0.0298	0.0078	0.1055	0.4037	0.1707	0.3274	0.2644	0.1630	0.0291	0.7254	0.2341	0.2857	0.300	0.1108	0.2729	-14.2295	1.8051	2025-02-24-0801	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.470434
28	ransac_regressor	-7.8132	-7.8384	-24.1914	-7.6361	1.2185	0.2594	0.5383	0.4586	1.4847	0.0673	1.1931	0.3482	0.2954	0.300	-0.1727	0.8236	-18.2750	1.9430	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.807614
29	theil_sen_regressor	-4.1146	-4.1293	-5.1473	-3.9803	0.9283	0.2488	0.4789	0.3975	0.8617	0.0619	1.0613	0.3354	0.2711	0.300	-0.1505	0.6275	-14.8497	1.9055	2025-02-24-0801	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 48 second(s)	0:00:48.660698
30	svm_regressor_linear_1	-0.0516	-0.0546	-0.0061	-0.0285	0.4209	0.1573	0.3002	0.2220	0.1772	0.0248	0.6653	0.2118	0.2399	0.300	-0.0624	0.2845	-11.4650	1.9284	2025-02-24-0801	2025-02-24-0801	0 day(s), 0 hour(s), 0 minute(s), 5 second(s)	0:00:05.020779
31	elastic_regressor_cv_2	-0.0001	-0.0029	-0.0020	0.0000	0.4105	0.1653	0.3274	0.2487	0.1685	0.0273	0.7257	0.2253	0.2864	0.300	-0.0030	0.2774	-11.9391	1.9357	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.882527
32	elasticnet_regressor	-0.0001	-0.0029	-0.0020	0.0000	0.4105	0.1653	0.3274	0.2487	0.1685	0.0273	0.7257	0.2253	0.2864	0.300	-0.0030	0.2774	-11.9391	1.9357	2025-02-24-0757	2025-02-24-0757	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.187823
33	lasso_regressor_cv	-0.0001	-0.0029	-0.0020	0.0000	0.4105	0.1653	0.3274	0.2487	0.1685	0.0273	0.7257	0.2253	0.2864	0.300	-0.0030	0.2774	-11.9391	1.9357	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.558145
34	lasso_regressor	-0.0001	-0.0029	-0.0020	0.0000	0.4105	0.1653	0.3274	0.2487	0.1685	0.0273	0.7257	0.2253	0.2864	0.300	-0.0030	0.2774	-11.9391	1.9357	2025-02-24-0800	2025-02-24-0800	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.191016

	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	extra_trees_regressor	0.7349	0.7343	0.7459	0.7354	0.2064	0.0832	0.1456	0.1088	0.0426	0.0069	0.3141	0.1048	0.1030	0.261	0.0083	0.1463	-12.2596	1.9991	2025-02-24-0806	2025-02-24-0807	0 day(s), 0 hour(s), 0 minute(s), 44 second(s)	0:00:44.096760
1	random_forest_regressor	0.7327	0.7321	0.7371	0.7330	0.2072	0.0838	0.1488	0.1112	0.0429	0.0070	0.3210	0.1073	0.1084	0.261	0.0062	0.1469	-12.2042	2.0024	2025-02-24-0811	2025-02-24-0812	0 day(s), 0 hour(s), 1 minute(s), 40 second(s)	0:01:40.381868
2	random_forest_regressor_2	0.7181	0.7175	0.7265	0.7183	0.2128	0.0858	0.1536	0.1146	0.0453	0.0074	0.3314	0.1105	0.1127	0.261	0.0053	0.1509	-12.1604	2.0085	2025-02-24-0812	2025-02-24-0813	0 day(s), 0 hour(s), 1 minute(s), 13 second(s)	0:01:13.746002
3	catboost_regressor	0.7080	0.7073	0.7195	0.7081	0.2166	0.0876	0.1597	0.1191	0.0469	0.0077	0.3446	0.1153	0.1227	0.261	0.0038	0.1536	-12.1391	2.0195	2025-02-24-0815	2025-02-24-0816	0 day(s), 0 hour(s), 0 minute(s), 59 second(s)	0:00:59.656978
4	xgb_regressor_2	0.6937	0.6930	0.7049	0.6939	0.2218	0.0896	0.1635	0.1216	0.0492	0.0080	0.3527	0.1176	0.1215	0.261	0.0052	0.1573	-12.2109	2.0202	2025-02-24-0810	2025-02-24-0811	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.743354
5	xgb_regressor_1	0.6937	0.6930	0.7049	0.6939	0.2218	0.0896	0.1635	0.1216	0.0492	0.0080	0.3527	0.1176	0.1215	0.261	0.0052	0.1573	-12.2109	2.0202	2025-02-24-0810	2025-02-24-0810	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.035324
6	lgbm_regressor	0.6758	0.6750	0.6907	0.6758	0.2282	0.0923	0.1699	0.1271	0.0521	0.0085	0.3666	0.1224	0.1305	0.261	0.0037	0.1618	-12.0730	2.0058	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.622421
7	hist_gradient_boosting_regressor	0.6729	0.6722	0.6876	0.6731	0.2292	0.0925	0.1707	0.1274	0.0525	0.0086	0.3682	0.1228	0.1289	0.261	0.0044	0.1625	-12.0961	1.9992	2025-02-24-0808	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 4 second(s)	0:00:04.848748
8	knn_regressor_2	0.6479	0.6471	0.6682	0.6496	0.2378	0.0954	0.1666	0.1247	0.0566	0.0091	0.3595	0.1184	0.1160	0.261	0.0166	0.1686	-12.4340	1.9598	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.773485
9	knn_regressor_1	0.6479	0.6471	0.6682	0.6496	0.2378	0.0954	0.1666	0.1247	0.0566	0.0091	0.3595	0.1184	0.1160	0.261	0.0166	0.1686	-12.4340	1.9598	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.506664
10	gradient_boosting_regressor_1	0.5646	0.5637	0.5861	0.5646	0.2645	0.1066	0.2021	0.1509	0.0699	0.0114	0.4361	0.1450	0.1631	0.261	0.0008	0.1875	-11.8978	2.0052	2025-02-24-0807	2025-02-24-0808	0 day(s), 0 hour(s), 0 minute(s), 37 second(s)	0:00:37.007226
11	mlp_regressor_2	0.5319	0.5308	0.5472	0.5322	0.2743	0.1090	0.2048	0.1492	0.0752	0.0119	0.4418	0.1445	0.1592	0.261	-0.0072	0.1944	-11.8433	2.0298	2025-02-24-0813	2025-02-24-0815	0 day(s), 0 hour(s), 1 minute(s), 9 second(s)	0:01:09.033382
12	decision_tree_regressor	0.5162	0.5151	0.4908	0.5162	0.2788	0.1116	0.1885	0.1383	0.0777	0.0125	0.4067	0.1346	0.1200	0.261	-0.0000	0.1977	-12.3974	1.9899	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.130242
13	mlp_regressor_1	0.5135	0.5124	0.5373	0.5158	0.2796	0.1119	0.2118	0.1579	0.0782	0.0125	0.4569	0.1496	0.1651	0.261	0.0190	0.1982	-12.2498	2.0266	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 46 second(s)	0:00:46.254962
14	decision_tree_pruned_regressor	0.4763	0.4751	0.4979	0.4764	0.2901	0.1169	0.2216	0.1657	0.0841	0.0137	0.4780	0.1582	0.1812	0.261	0.0037	0.2057	-11.9500	1.9980	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.929447
15	ada_boost_regressor	0.3454	0.3440	0.3866	0.4318	0.3243	0.1335	0.2662	0.2128	0.1052	0.0178	0.5744	0.1893	0.2394	0.261	0.1178	0.2299	-14.1046	1.7286	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 8 second(s)	0:00:08.184067
16	ard_regressor	0.2969	0.2954	0.3202	0.2969	0.3361	0.1364	0.2632	0.1996	0.1130	0.0186	0.5677	0.1882	0.2204	0.261	-0.0017	0.2383	-11.6699	2.0406	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.468076
17	elasticnet_regressor_cv	0.2968	0.2953	0.3203	0.2968	0.3361	0.1364	0.2633	0.1997	0.1130	0.0186	0.5679	0.1884	0.2206	0.261	-0.0017	0.2383	-11.6688	2.0406	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 3 second(s)	0:00:03.851112
18	linear_regressor	0.2967	0.2952	0.3204	0.2967	0.3361	0.1364	0.2632	0.1995	0.1130	0.0186	0.5678	0.1882	0.2207	0.261	-0.0017	0.2383	-11.6715	2.0406	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.411814
19	ridge_regressor	0.2967	0.2952	0.3204	0.2967	0.3361	0.1364	0.2632	0.1995	0.1130	0.0186	0.5678	0.1882	0.2207	0.261	-0.0017	0.2383	-11.6714	2.0406	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.424298
20	bayesian_ridge_regressor	0.2967	0.2951	0.3204	0.2967	0.3362	0.1364	0.2632	0.1996	0.1130	0.0186	0.5679	0.1883	0.2206	0.261	-0.0017	0.2383	-11.6704	2.0405	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.275674
21	ridge_regressor_cv	0.2967	0.2952	0.3204	0.2967	0.3361	0.1364	0.2632	0.1995	0.1130	0.0186	0.5678	0.1882	0.2207	0.261	-0.0017	0.2383	-11.6713	2.0406	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.724564
22	ridge_regressor_2	0.2967	0.2952	0.3204	0.2967	0.3361	0.1364	0.2632	0.1995	0.1130	0.0186	0.5678	0.1882	0.2207	0.261	-0.0017	0.2383	-11.6714	2.0406	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.220097
23	huber_regressor	0.2926	0.2911	0.3158	0.2965	0.3371	0.1358	0.2609	0.1943	0.1137	0.0185	0.5627	0.1861	0.2162	0.261	-0.0251	0.2390	-11.2414	2.0326	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.131763
24	svm_regressor_linear_1	0.2856	0.2841	0.3091	0.2932	0.3388	0.1358	0.2602	0.1920	0.1148	0.0184	0.5614	0.1850	0.2111	0.261	-0.0348	0.2402	-11.0440	2.0254	2025-02-24-0809	2025-02-24-0810	0 day(s), 0 hour(s), 0 minute(s), 2 second(s)	0:00:02.808430
25	elasticnet_regressor_2	0.2731	0.2715	0.2910	0.2731	0.3418	0.1386	0.2686	0.2058	0.1168	0.0192	0.5795	0.1926	0.2295	0.261	-0.0034	0.2423	-11.4910	2.0583	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.341761
26	lasso_regressor_2	0.2207	0.2190	0.2331	0.2208	0.3538	0.1438	0.2795	0.2159	0.1252	0.0207	0.6029	0.2007	0.2408	0.261	-0.0041	0.2509	-11.3901	2.0752	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.268478
27	ransac_regressor	-14.2797	-14.3133	-0.9244	-14.2456	1.5668	0.3978	0.7968	0.7061	2.4550	0.1582	1.7198	0.4811	0.3150	0.261	-0.0741	1.1109	-25.3650	2.0451	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.844620
28	theil_sen_regressor	-0.4904	-0.4936	-0.5668	-0.4893	0.4893	0.2085	0.3741	0.2957	0.2395	0.0435	0.8072	0.2838	0.2941	0.261	0.0133	0.3469	-12.8038	2.0139	2025-02-24-0810	2025-02-24-0810	0 day(s), 0 hour(s), 0 minute(s), 54 second(s)	0:00:54.066371
29	svm_regressor_linear_2	-0.1470	-0.1495	-0.0950	-0.0947	0.4293	0.2020	0.3342	0.2580	0.1843	0.0408	0.7211	0.2853	0.2659	0.261	-0.0916	0.3044	-10.9357	1.8943	2025-02-24-0810	2025-02-24-0810	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.234122
30	gradient_boosting_regressor_2	-0.0002	-0.0024	0.0349	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	0.0271	0.6957	0.2316	0.2849	0.261	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0808	2025-02-24-0808	0 day(s), 0 hour(s), 0 minute(s), 35 second(s)	0:00:35.071838
31	elastic_regressor_cv_2	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	0.0271	0.6957	0.2316	0.2849	0.261	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 1 second(s)	0:00:01.680076
32	elasticnet_regressor	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	0.0271	0.6957	0.2316	0.2849	0.261	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0806	2025-02-24-0806	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.291471
33	lasso_regressor_cv	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	0.0271	0.6957	0.2316	0.2849	0.261	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.829974
34	lasso_regressor	-0.0002	-0.0024	-0.0007	0.0000	0.4009	0.1646	0.3225	0.2531	0.1607	0.0271	0.6957	0.2316	0.2849	0.261	-0.0056	0.2842	-11.2838	2.0870	2025-02-24-0809	2025-02-24-0809	0 day(s), 0 hour(s), 0 minute(s), 0 second(s)	0:00:00.239740

	outlier_transformation	model_name	r2	adjusted_r2	cross_validated_r2	explained_variance_score	rmse	rmsle	mae	mape	mse	msle	mase	smape	med_ae	huber_loss	mbd	cv_rmse	r2_residuals	durbin_watson_stat	start_time	end_time	analysis_duration_text	analysis_duration
0	zscore	extra_trees_regressor	0.7732	0.7727	0.7618	0.7735	0.1878	0.0751	0.1306	0.0969	0.0353	0.0056	0.2956	0.0933	0.0882	0.2957	0.0065	0.1340	-12.4620	2.0112	2025-02-24-0739	2025-02-24-0740	0 day(s), 0 hour(s), 1 minute(s), 0 second(s)	0:01:00.843443
1	before_transformation	extra_trees_regressor	0.7657	0.7652	0.7694	0.7663	0.1940	0.0766	0.1346	0.0986	0.0376	0.0059	0.2904	0.0947	0.0911	0.3139	0.0098	0.1376	-12.2819	2.0139	2025-02-24-0715	2025-02-24-0716	0 day(s), 0 hour(s), 0 minute(s), 43 second(s)	0:00:43.667105
2	clipped_iqr	extra_trees_regressor	0.7349	0.7343	0.7459	0.7354	0.2064	0.0832	0.1456	0.1088	0.0426	0.0069	0.3141	0.1048	0.1030	0.2609	0.0083	0.1463	-12.2596	1.9991	2025-02-24-0806	2025-02-24-0807	0 day(s), 0 hour(s), 0 minute(s), 44 second(s)	0:00:44.096760
3	dbscan	extra_trees_regressor	0.7288	0.7280	0.7383	0.7288	0.2138	0.0820	0.1463	0.1015	0.0457	0.0067	0.3242	0.0986	0.0980	0.2997	0.0015	0.1445	-12.7016	2.0774	2025-02-24-0757	2025-02-24-0758	0 day(s), 0 hour(s), 0 minute(s), 45 second(s)	0:00:45.592145
4	iqr	extra_trees_regressor	0.6825	0.6814	0.6787	0.6842	0.2174	0.0828	0.1526	0.1031	0.0473	0.0068	0.3540	0.0991	0.1086	0.2696	0.0158	0.1411	-15.9005	2.0047	2025-02-24-0730	2025-02-24-0731	0 day(s), 0 hour(s), 0 minute(s), 41 second(s)	0:00:41.630663
5	modified_zscore_mad	extra_trees_regressor	0.5823	0.5804	0.5754	0.5825	0.2402	0.0880	0.1677	0.1037	0.0577	0.0078	0.4083	0.1020	0.1125	0.2752	-0.0054	0.1454	-19.1720	1.9675	2025-02-24-0751	2025-02-24-0751	0 day(s), 0 hour(s), 0 minute(s), 19 second(s)	0:00:19.275262

This notebook is primarily based on references, code, and notes from:¶

Additional References:¶

configure dependencies¶

imports¶

Helper Functions¶

Get Time¶

Calculate Time Duration¶

Display Sum NaN Function¶

Check Unique Cols¶

Feature Importance¶

Display BoxPlot of Features¶

Separate Function to display Histogram and KDE Plot¶

Display Histogram and KDE Plot¶

Create Holdout and Standardize the Dataset¶

Display Dataframe as HTML Styled Widget¶

Import Data¶

View Columns¶

Rename Cols¶

Drop Column¶

Dataset Characteristics¶

Boston House Prices¶

Save updated dataset¶

PPrint DF¶

View Sample of the Dataset¶

View df¶

Check info¶

Assign numerical, non-numerical cols and dependent var name¶

Check unique Counts¶

Describe the df¶

Set df of Features¶

View Population Variation¶

Time Stamp¶

Exploratory Data Analysis (EDA)

Empirically Analyze Feature Distributions¶

View Feature Distributions - Univariate Analysis¶

Large tails of kde suggest significant outliers that we should deal with later¶

View Boxplot of Features - Understand centrality of our data¶

Bivariate Analysis

Correlation Analysis¶

Scatterplot¶

linear model plot --> scatter plot with a regression line¶

Violinplot¶

Pairplot¶

Plot Error Bar of Distribution of Features¶

Bifurcate Independent and Dependent Variables¶

Create Hold-Out (Train and Test Set)¶

Standardize the Dataset¶

Timestamp¶

Visualize Standardized Dataset ¶

View Boxplot of Standardized Features¶

Observations: Our data is still heavily affected by outliers. We will proceed first with modeling the current data to measure performance metrics, then we will adjust outliers and re-model to empirically measure improvements with handling outliers in the dataset¶

note: this still looks like there are heavy outliers affecting the datasets. We will proceed with modeling the current data, and then return and repeat after we handle outliers and evaluate if our performance improves¶

Mean Absolute Error (MAE):

$\frac 1n\sum_{i=1}^n|y_i-\hat{y}_i|$ ¶

Mean Squared Error (MSE):¶

$\frac 1n\sum_{i=1}^n(y_i-\hat{y}_i)^2$ ¶

Root Mean Squared Error (RMSE):¶

$\sqrt{\frac 1n\sum_{i=1}^n(y_i-\hat{y}_i)^2}$ ¶

R-Squared (R2) (Coefficient of Determination):

$R^2 = 1- \frac {SSR}{SST}$ ¶

Adjusted R-Squared :

$R^2_{adj.} = 1 - (1-R^2)*\frac{n-1}{n-p-1}$ ¶

Cross-Validated R-Squared ($ R^2_{\text{cv}} $):

$R^2_{\text{cv}} = \frac{1}{K} \sum_{i=1}^{K} R^2_{\text{fold}_i}$ ¶

Explained Variance Score (EVS):

$EVS = 1 - \frac{Var(Y_{\text{TEST}} - y_{\text{pred}})}{Var(Y_{\text{TEST}})}$ ¶

Root Mean Squared Logarithmic Error (RMSLE):¶

$\sqrt{\frac 1n\sum_{i=1}^n(\log(1 + y_i) - \log(1 + \hat{y}_i))^2}$ ¶

Mean Squared Logarithmic Error (MSLE):¶

$\frac 1n\sum_{i=1}^n(\log(1 + y_i) - \log(1 + \hat{y}_i))^2$ ¶

Huber Loss:

$ L(y, ŷ) = \begin{cases}¶

Mean Bias Deviation (MBD):¶

$\frac 1n\sum_{i=1}^n(\hat{y}_i - y_i)$ ¶

Coefficient of Variation of RMSE (CVRMSE):¶

$\frac{RMSE}{\bar{y}} * 100$ ¶

R-squared of Residuals:¶

Durbin-Watson Statistic (DW):¶

$DW = \frac{\sum_{t=2}^{n} (e_t - e_{t-1})^2}{\sum_{t=1}^{n} e_t^2}$ ¶

Mean Absolute Percentage Error (MAPE):¶

**$R^2_{adj.} = 1 - (1-R^2)*\frac{n-1}{n-p-1}$** ¶

**$\frac{RMSE}{\bar{y}} * 100$** ¶

**Create new method to clip outliers to 1.5*IQR Value** ¶